"/\:.5"j&
LBL-140O
A INORGANIC MATERIALS RESEARCH DIVISION Annual Report 1972
THIS DOCUMENT CONFIRMED'AS • UNCLASSIFIED ' ^f^' DiV*SlQN (^CLASSIFICATION ' Civ-" 1ST H r Lawrence o Berkeley Laboratory ' ,& University of California Berkeley 8
r.ffilREUT'O; 0A TH'S DCSU.ACFJT iS i mm SnSialliirgkni Enginesfin; ^w? ~fc<£t~-
Cheml: Solid-Slots Chtmicol l#r«inic Physic* tso Enginoer'rna Er gTneerinjj LBL-1400 UC-2 General,Miscellaneous, and Progross Reports TID-4500 (59th Ed.)
INORGANIC MATERIALS RESEARCH DIVISION ANNUAL REPORT, 1972
April 1973
-NOTICE- This report was prepared as en account of work sponsored by the United States Government. Neither the United States nor the United States Atomic Energy Commission, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for *k* acaeacy, com- pletcness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not Infringe privately owned rights.
Lawrence Berkeley Laboratory University of California Berkeley, California
&l»U.ONOPTH1SDOCUMENT1SONU^ Work done under U.S. Atomic Energy Commission Contract No. W-7405-eng-48
Printed in the United States of America Available from National Technical Informotion Service U.S. Department of Commerce 5285 Port Royal Road Springfield, Virginia 22151 Price: Printed Copy $*8»; Microfiche $0.95 -iii-
INORGANIC MATERIALS RESEARCH DIVISION ANNUAL REPORT. 1972*
Contents
GENERAL INTRODUCTION ix
I. CHEMISTRY A. Inorganic Chemistry William h. Jolly, Principal Investigator 1. X-Ray Photoelectron Spectroscopy 1 2. Boron Hydride Chemistry 4 3. Sulfur-Nitrogen Chemistry 5 4. The Prediction of Gas-Phase lewis Acid-Base Dissociation Heats with an Empirical Four-Parameter Equation 6 5. Research Plans for Calendar Year 1973 7 6. 1972 Publications and Reports , 7 Robert E. Connick, Principal Investigator 1. Oxygen-17 NMR Study of the Rate of Water Exchange from Partially Complexed Nickel Ion 10 2. 170-NMR Studies of Water Molecules in the Second Coordination Sphere of Cr3+ 10 3. Computer Simulation of Solvent Exchange from the First Coordination Spheres of (.fetal Ions 10 4. Raman Spectrum of Triiodide Ion 10 5. Research Plans for Calendar Year 1973 10 6. 1972 Publications and Reports 11 Neil Bartlettj Principal Investigator 1. Noble-Gas and Halogen Chemistry 12 2. Fluorides, Oxyfluorides, and Oxides of the Transition Mstals 15 3. Research Plans for Calendar Year 1973 16 4. 1972 Publications and Reports 17 B. Solid State Chemistry and Physics Norman E. Phillips, Principal Investigator 1. Heat Capacity of SmS at 0 and IS kbar 19 2. Heat Capacity of PrCu? 19 3. The Heat Capacity of Dilute Cu Mi Alloys 19 4. The Low-Temperature Heat Capacity of Neodymiura 19 5. Research Plans for Calendar Year 1973 20 6. 1972 Publications and Reports 20 George Jura, Principal Investigator 1. Heat Capacities at High Pressures 21 2. Heat Capacity in Iron and Aluminum 22 3. Heat Capacity of Bismuth , 23 4. Research Plans for Calendar Year 1973 23 5. 1972 iublications and Reports 24 Gabov A. Sonbrjai, Principal Investigator 1. Low Energy Electron Diffraction Studies by Surfaces 25 2. Auger Electron Spectroscopy of Surfaces 27 3. Catalysis by Single-Crystal Surfaces 28 4. Molecular Beam Scattering by Surfaces 29 5. Research Plans for Calendar Year 1973 30 6. 1972 Publications and Reports 30 Lee F. Donaghey, Principal Investigator 1. Gas-Solid Processes 31 2. Crystallization Processes 32 3. Solid Electrolyte Studies 35 4. Research Plans for Calendar Year 1973 36 5. 1972 Publications and Reports 37
Preceding Annual Reports: UCRL-20500, LBL-425 -iv-
C. Electrochemistry Charles W. Tobiac, Principal Investigator 1. Fundamental Studies of Ionic Mass Transport 38 2. Electrolysis at High Current Densities 39 3. Electrochemistry in Nonaqueous Solvents 41 4. Research Plans for Calendar Year 1973 41 5. 1972 Publications and Reports 42 Rolf H. Muller, Principal Investigator 1. Optical Studies of Interfacial Phenomena 43 2. Research Plans for Calendar Year 1973 45 3. 1972 Publications and Reports 46 John Newman, Principal Investigator 1. Analysis of Porous Electrodes with Sparingly Soluble Reactants 47 2. Potential Distribution in Axisymnetric Mercury-Pool Electrolysis Cells at Limiting Current 47 3. Mass Transfer to a Non-Newtonian Falling Film 47 4. Mass Transfer in Cylindrical Couette Flow 47 5. Corrosion of an Iron Rotating Disk 48 6. Mass Transfer Coefficient in Porous Flow Through Electrodes 48 7. The Radially Dependent Convective Warburg Problem 48 8. A Composite Expansion for the Wake Just Behind a Flat Plate 48 9. Mass Transfer to a Plane Below a Rotating Disk at High Schmidt Numbers . . 49 10. The Transient Response of a Disk Electrode 49 11. Current Distribution at a Rotating Sphere Below the Limiting Current ... 49 12. Research Plans for Calendar Year 1973 50 13. 1972 Publications and Reports 50 D. Physical Chemistry Leo Brewer, Principal Investigator 1. Thermodynamic Properties of Third and Fourth Group Transition Metals with Platinum Group Metals 52 2. Application of the Regular Solution Theory co Metallic Solutions 52 3. Cptical Pumping in Rare Gas Matrices 52 4. Compilation of Ternary Phase Diagrams 52 5. Solubilities of Intermetallic Compounds in Iron 52 6. Thermodynamic Properties of Mo and its Compounds 52 7. Thermodynamic Properties and Phase Diagrams of the Alloy Systems of the Lanthanide and Actinide tetals 52 8. Research Plans for Calendar Year 1973 52 9. 1972 Publications and Reports 53 Charles B. Harris, Principal Investigator 1. Excited Triplet States in Molecules and Molecular Crystals 54 2. Coherent Energy Migration in Molecular Solids 54 3. Electron Spin Resonance Absorption in Coherent Triplet Exciton States ... 55 4. Exciton-Trap Interactions in the Coherent and Incoherent Limits 57 5. Phosphorescence Microwave Double Resonance Spectroscopy in Excited Triplet States 58 6. Breakdown of the Bom-Oppenheimer Approximation in Radiative Transitions . 59 7. Coherent Spin Phenomena in Molecular Excited States 59 8. Coherent Cptical Phenomena in Molecular Excited States 61 9. Research Plans for Calendar Year 1973 61 10. 1972 Publications and Reports 62 Alexander Pines, Principal Investigator 1. Nuclear Magnetic Double-Resonance 64 2. Nuclear Spin Thermodynanics and Multiple-Pulse Nuclear Magnetic Resonance 65 3. Research Plans for Calendar Year 1973 66 4. 1972 Publications and Reports 67 Ronald R. Eerm, Principal Investigator 1. Collision Dynamics of "Alkali-Like" Atoms 68 2. Crossed Beam Reactions of BaF2 with BCI3 68 3. Research Plans for Calendar Year 1973 69 4. 1972 Publications and Reports 69 -V-
Harcld S. Johnston, Principal Investigator Introduction 71 1. Experimental Studies 71 2. Theoretical Studies 73 3. Research Plans for Calendar Year 1973 73 4. 1972 Publications and Reports 74 Bruce U. Mohan, Principal Investigator 1. Recombination of Gaseous Ions 75 2. The 02+-C2l>2 Reaction: An Endoergic Stripping Process 75 + 3. Dynamics of the 0 -H2 Reaction. I: Reactive Scattering of 0+(4s3/2) at Relative Energies below 15 eV 76 4. Dynamics of the 0+-H2 Reaction. II: Reactive and Non-Reactive Scattering at Relative Energies above 15 eV 77 5. An Apparatus for the Study of Ion-Molecule Reactions at Low Relative Collision Energies 77 6. Research Plans for Calendar Year 1973 77 7. 1972 Publications and Reports 78 William H. hiiller, Principal Investigator 1. Semiclassical Quantization of Non-Separable Systems 79 2. Penning Ionization of Hydrogen Atoms by Metastable Helium Atoms 79 3. Analytic Continuation of Classical Mechanics for Classically Forbidden Processes in Inelastic and Reactive Scattering 80 4. Semiclassical Theory of Electronic Transitions in Low Energy Atomic and hblecular Collisions Involving Several Nuclear Degrees of Freedom ... 82 5. Research Plans for Calendar Year 1973 82 6. 1972 Publications and Reports 82 Rollie J. Myers, Principal Investigator 1. Molecular Field Corrections for Electron Paramagnetic
Resonance Spectroscopy + 84 2. Electron Paramagnetic Resonance Spectrum of Ti(H20)g 85 3. Proton Nuclear Magnetic Resonance on Solutions of Transition-Metal Ions . 86 4. Research Plans for Calendar Year 1973 87 5. 1972 Publications and Reports 87 Kenneth S. Pitzer, Principal Investigator 1. Solutions of Electrolytes 88 2. Other Topics 88 3. Research Plans for Calendar Year 1973 89 4. 1972 Publications and Reports 89 E. Nuclear Science Donald F. Olander, Principal Investigator 1. Laser Vaporization of Solids 90 2. A Modulated Molecular Beam Study of the Energy of Simple Gases Scattered from Pyrolytic Graphite 90 3. Molecular Beam Surface Reactions-Water Vapor on Graphite 90 4. Viscosity of Cesium up to 1600°C 91 5. Diffusion in Liquid Uranium 91 6. Oxygen Redistribution in Ceramic Nuclear Fuels 92 7. Research Plans for Calendar Year 1973 . 92 8. 1972 Publications and Reports 92 II. METALLURGY A. Crystal Imperfections Jack Washburn, Principal Investigator 1. Mechanism of Dipole Break-up 97 2. The Mechanism of Dislocation Loop Coarsening in MgO 97 3. Precipitation of Phosphorous in ion Implanted Silicon 98 4. Memory State Filaments in As-Te-I Amorphous Semiconductors 100 5. Field Ion Microscope Observations of Low Tenperature Ion Bombardment Damage in Iridium 100 6. Research Plans for Calendar Year 1973 101 7. 1972 Publications and Reports 101 -vi-
B. Tlieoretical Metallurgy John W. Monn-s, Jr., Principal Investigator 1. Thermodynamics and Phase Stability 103 2. Mechanical Properties 104 3. Researdi Plans for Calendar Year 1973 106 4. 1972 Publications and Reports 106 C. Relation of Microstructure and Properties Gareth Thomas, Principal Investigator Introduction 108 1. Structure, Strength and Toughness of Fe-Cr-C Martensitic Steels 108 2. Martensitic Transformation and Strengthening of a Spinodal Fe/Ni/Cu Alloy 109 3. The Tempering of Martensite in an Iron 1.55 Nitrogen 111 4. Intergranular Fracture of Aged Cu-Ni-Fe Alloys 112 5. Short Range Order in Ni-Mo, Au-Cr, Au-V, and Au-Mn Alloys 112 K. Hign Voltage Electron Microscopy Gareth Thomas, Principal Investigator Introduction 114 1. Transndssion Electron Microscopy at 2.5 MeV 114 2. Defect Contrast and Resolution in High Voltage Electron Microscopy .... 115 3. Ksjn Voltage Electron Microscopic Studies of Minerals 116 4. -Jji Angle Tilting Stage for the 650 kV Hitachi Microscope 119 5. Litliium Ferrite 119 6. Other Research Projects in Progress ... 119 7. Research Plans for Calendar Year 1973 120 8. 1972 Publications and Reports 120 E. Alloy Design Victor F. Zackay and Earl R. Parker, Principal Investigators Introduction 122 1. The Fracture Toughness of Ultra-High Strength Alloy Steels 122 2. Structure and Fracture Toughness of Secondary Hardening Steels 125 3. Relationship Between Fatigue Crack Growth Rate and Microstructure in a Secondary Hardening Steel 127 4. The Influence of Cooling Rate in a Jominy Test on Toughness of Steels . . 129 5. Design of Interstitial-Free Cryogenic Alloys 132 6. Mechanical Properties of the Cryogenic Alloy Fe-12Ni-0.25Ti 134 7. Structure and Mechanical Properties of Non-Carbon-Containing Ferritic Alloys at Room and Elevated Temperatures 135 8. Structure and Mechanical Properties of TRIP Steels Processed by Deformation and Thermal Cycling 137 9. Design of High Manganese Metastable Austenitic Steels 138 10. Isothermal Studies of Phase Transformations in Steels 139 11. Researdi Plans for Calendar Year 1973 140 12. 1972 Publications and Reports 140 F. High Field Superconductivity Earl R. Parker, Victor F. Zackay and Milton R. Piakus, Principal Investigators 1. Introduction 142 2. Evaluation of Optimal Parameters for Powder-Rolled Tape 143 3. A New Process for Making Stabilized Superconducting Tape 145 4. The Niobium-Molybdenum Nitride System 145 5. Research Plans for Calendar Ye^.r 1973 X46 6. 1972 Publications and Reports 146 G. Powder Metallurgy Milton R. Pickus, Principal Investigator 1. Introduction 147 2. Research Plans for Calendar Year 1973 j. ? H. Thermodynamics of Metal Systems Ralph R. Hultgren, Principal Investigator 1. Evaluation of Thermodynamic Data 148 2. Metals Handbook Phase Diagrams 148 3. Thermodynamic Evaluation and Compilation of Metal Alloy Systems 148 4. 1972 Publications and Reports . 148 -vii-
I. Advanced Materials Robert H. Bragg, Principal Investigator 1. Structure and Properties of Sone Advanced Materials 149 2. Structure of Glassy Carbon 149 3. Small-Angle Scattering by Oriented Ellipsoids 149 4. Unidirectional Solidification of AI-Q1AI2 Eutectic 149 5. Other Work in Progress ISO 6. Research Plans for Calendar Year 1973 ISO 7. 1972 Publications and Reports 150 III. CERAMIC SCIENCE A. High Temperature Reactons titan W. Searcy, Principal Investigator X. The Kinetics of Vaporization of Barium Sulfate 155 2. The Fragmentation of Dimers of Metal Trihalides by Electron Impact .... 155 3. The Mechanism of ZnO Single-Crystal Vaporisation 156 4. Scattering in a Molecular Beam as a Result of Intrabeam Collisions .... 156 5. Gaseous Thallium (I) Mstaborate and Thallium (I) Aluminum Fluoride .... 156
6. The Zeeman Effect in Se2 157 7. The Conditions of Chemical Equilibrium 157 8. The Relationship Between Forward and Reverse Reaction Rates under Nonequilibrium Conditions 157 9. Research Plans for Calendar Year 1973 158 10. 1972 Publications and Reports 158 B. Microstructure and Behavior of Ceramic Materials: Glass and Ceramic-Metal Systems Joseph A. Pask, Principal Investigator 1. Stable and Metastable Equilibria in the Silica-Alumina System 160 2. Diffusion and Reaction Studies in the System Al203-Si02 160 3. Diffusion Studies in the S1O2-AI2O3 System in the Temperature Range of 1800 to 2000°C 161 4. Densities of S1O2-AI2O3 Melts 162 5. Permeation of Silicate Liquids into Sintered Magnesia Compacts 163 6. Polycrystalline Spinel from Powders Prepared by Freeze Drying Technique 163 7. Liquid Phase Sintering Studies 164 8. Impact Fracture of Brittle Transparent Materials 164 9. Mechanical Behavior of MgO Single Crystals 164 10. Effect of Grain Boundary Character on High Temperature Mechanical Behavior of PolycrystcJAine MgO 165 11. Control of Grain Boundary Character in Polycrystalline MgO 166 12. Interfacial Reactions and Wetting Behavior of Glass-Iron Systems 166 13. Effect of Compositions on Glass-Metal Interface Reactions and Adherence 167
14. Electrochemical Studies of the Fen asO-Na20-Si02 Glass System 167 15. Research Plans for Calendar Year 1973 167 16. 1972 Publications and Reports 168 C. Relation of Microstructure to Properties in Ceramics Richard M. Sulrath, Principal Investigator 1. Hot Stage SEM Studies of the Initial Stage Sintering of Copper and Nickel 169 2. Sintering Ceramic Oxides 170 3. Intrinsic and Extrinsic Nonstoichiometry in the Lead Zirconate-Titanate System 171 4. Processing Parameters and the Properties of Lithium Ferrite ... 177 5. Diffusion of Gases in Glass 179 6. Research Plans for Calendar Year 1973 181 7. 1972 Publications and Reports 181 -viii-
IV. SOLID STATE PHYSICS A. Theoretical Solid State Physics Marvin L. Cohen, Principal Investigator Introduction 185 1. Semiconductors and Insulators 185 2. Metals 186 3. Superconductivity 186 4. Research Plans for Calendar Year 1973 186 5. 1572 Publications and Reports 186 B. Far Infrared Properties of Solids Paul L. Richards, Principal Investigator 1. Resonance Modes in Antiferromagnetic (Co, Mn)F2 and (Fe, 188 2. Balloon-Bome Far Infrared Spectrometer 189 3. Ideal Performance from Point Contact Josephsm Junctions 189 4. Instrumentation for Fourier Transform Spectroscopy -.-... 190 5. Far Infrared Studies of y-Irradiated Ruby 190 6. Generation of Tunable Far Infrared Radiation by Mixing Two Dye Lasers . . . 191 7. Nonlinear Optical Spectroscopy of x' ' in LiNbO? 192 8. Research Plans for Calendar Year 1973 192 9. 1972 Publications and Reports '. . 192 C. Superconductivity Gene I. Roahlin, Principal Investigator 1. Shunted Josephson Junctions: Experiments and Model 194 2. Behavior of Weak-Link Devices Near the Junction-Like to Bulk-Like Transition 194 3. Indium-Barrier-Indium Tunneling Junctions 194 4. Characterization of Tunneling Barriers 195 5. Observability of Quasiparticle-Pair Interference Current in Superconducting Weak Links 195 6. Research Plans for Calendar Year 1973 196 7. 1972 Publications and Reports 196 John Clarke, Principal Investigator 1. Non-Equilibrium Effects in Superconductors 197 2. Electron Focusing in Bismuth: A New Method of Measuring the Caliper of a Fermi Surface 197 3. An Improved DC Squid 198 4. A Josephson Junction Bolometer 199 5. Resistance of SNS Sandwiches at Very Low Temperatures 199 6. A Search for Superconductivity in Degenerate Semiconductors 200 7. 1/f Noise in Thin Films 201 8. Quantum Transitions in a DC Squid 201 9. Research Plans for Calendar Year 1973 202 10. 1972 Publications and Reports 203 D. Experimental Solid State Physics and Quantum Electronics J. Bon Shen, Principal Investigator 1. Resonance Raman Scattering in InSb 205 2. Resonance Raman Scattering and Luminescence in CU2O 205 3. Raman and Rayleigh Scattering in Liquid Crystals 205 4. Rayleigh-Wing Scattering Near the Nematic-Isotropic Transition 205 5. Wavelength-Modulation Spectroscopy of Solids 206 6. Dispersion of Nonlinear Optical Susceptibilities 206 7. Second-Harmonic Generation from a Solid-State Plasma 206 8. Far Infrared Generation by Optical Mixing 206 9. Coherent Phonon Generation by Optical Mixing 207 10. Spectral Broadening in the Self-Focused Light 207 11. Research Plans for Calendar Year 1973 207 12. 1972 Publications and Reports 207 APPENDICES IMRD STAFF FOR 1972 211 VISITING SCIENTISTS 216 DEGREES AWARDED IN 1972 217 1972 RESEARCH SEMINARS 219 GENERAL INTRODUCTION
Leo Brewer, Division Bead Victor F. Zackay, Associate Head Rolf H. Mutter, Assistant to Division Head
During 1972, 21 Ph.D., one Doctor of Engi administration. Professor Victor Zackay is neering and 16 M.S. degrees were awarded to Associate Dean and Professor Joseph Pask is BIRD students and 189 journal articles or book an Assistant Dean of the College of Engineer chapters were published including 18 articles ing. Professor Rollie Myers is Assistant Dean based on work supported by agencies other than of the College of Chemistry, and Professor the Atomic Energy Comjnission--D.O.T., N.S.F., Charles Tobias was Acting Director of the Art O.N.P.., U. S. Army, and A.F.O.S.R. The level Museun for the Fall quarter and completed his of support by other agencies now amounts to term as Chairman of the Department of Chemical approximately 11% of the total IMRD budget. Engineering on July 1, 1972. The number of postdoctorals is at an all-time high of 45 due to the Presidential Internships Several of the division's investigators and Public Employment Program (PEP) fellows received awards and honors in 1972: and an increased number of sabbatical visitors Professor Neil Bartlett--honorary Doctorate offsetting the reduction in funds available of Science degree from Colby College, Maine. for postdoctoral positions. The number of Professor Earl Parker--Gold Medal from the graduate students has remained roughly level American Society for Metals, and the at about 180 for the past five years. The Distinguished Teaching Award by the Department availability of positions for our graduates of Materials Science and Engineering. at both the M.S. and Ph.D. levels has dra Professor Charles Tobias--the Edward matically improved this last year and we Goodrich Acheson Award from the Electro again have the situation of students with chemical Society. many job offers. Several investigators are on sabbatical Ralph Hultgren retired officially on July leave: 1, 1972, but the completion of Vol. I. Professor Marvin Cohen--University of Paris, Selected Values of the Thermodynamic Properties France (September through June). of the Elements and Vol. II. Selected Values Professor Donald Olander—The Institute of of the Thermodynamic Properties of Binary Atomic and Molecular Physics, Amsterdam, Alloys, took all of his time through the end Netherlands (September through June) of the year. He will continue to work part Professor Earl Parker—In residence time on the maintenance of an up-to-date (September through June) bibliography for the thermodynamic properties Professor Yuen-Ron Shen--Harvard University of metals and alloys. (September through January) (University of Paris, France, remainder of leave) Professor Alexander Pines has become associated with the division during this year. He came from Massachusetts Institute of During the past several years, several Technology where he received his Ph.D. degree principal investigators in the Metallurgy in 1972 under the direction of Professor group, have successfully applied fundamental J. S. Waugh. His previous education had been principles of alloy theory, dislocation at the Hebrew University, Jerusalem. He has theory, theories of nucleation and growth, been very productive with 22 publications to and knowledge of the effects of processing to date and has started the development of a produce a variety of new and potentially use nuclear magnetic-double-resonance program ful materials. Some specific results are the designed to increase sensitivity sufficiently new TRIP steels, new superconducting wire to study structural and dynamical features of made from powdered niobium and liquid tin impurities in solids and many other NMR containing filaments of Nb3Sn, low alloy measurements not yet accessible. ultra high strength steels with exceptionally high fracture toughness, a strong high tem As in the past, a number of Inorganic Ma perature material, and a cryogenic alloy terials Research Division principal investiga more resistant to brittle fracture than any tors are participating in the University commercially available product. The success of these investigators en of a simple fluid has been obtained through couraged them to establish within the frame use of a multipolar model, allowing earlier work cf IMRD and LBL, a Center for the Design mechanical models to be reconciled. of Alloys to provide a focus on the efforts (J. W. Morris, Jr., p. 103) within BSD to create new and useful alloys through the application of basic principles Alloy design to apply spinodal transforma of materials science. The principal in tion in conjunction with the martensitic vestigators who participated in the estab transformation has produced high strength lishment of the Center were victor F. Zackay, dispersions similar to ausforming without use Earl R. Parker and Milton R. Pickus. Their of plastic deformation. research efforts are devoted almost entirely (G. Thomas, p. 108) to the design of new alloys with superior properties. Shortly before this Annual Review Following a study of the microstructural went to press, two more IMRD principal investi and compositional features that control 'die gators with an interest in alloy design joined toughness and strength at cryogenic teii}Kra- tile center—they are Gareth Thomas and J. W. tures, interstitial-free iron based alloys Morris, Jr. were designed, containing 12% nickel and varying amounts of titanium. Charpy impact The Sixth Annual IMRD Review meeting was energy values of ISO ft lbs and yield strength held on February 2 and 3, 1972. Reviewers of ISO ksi were attained at -196°C in the were: Professor C. S. Barrett, University of alloy Fe-12Ni-0.STi by heat treatment in a Denver; Professor Bneritus, J. Chipmnn, two phase a+Y region. In further studies Massachusetts Institute of Technology; Prof (by Professor J. W. Morris), the titanium essor W. Klemperer, Harvard University; content was decreased to 0.254 and the alloy Professor N. Nachtrieb, The University of had a Charpy impact value of 172 ft lbs and Chicago; Professor R. W. Parry, University of yield strength of 185 ksi at 6°K. Utah; and Professor S. V. Radcliffe, Case CV. F. Zackay and E. R. Parker, p. 122) Western Reserve University. Work on optimization of the process param eters of the (IMRD developed) powder rolled superconducting tape has resulted in a six StME HIGHLIGHTS OF THE 1972 RESEARCH PROGRAM fold increase in current carrying capacity. It has been established that niobium hydride The novel oxidation state of gold AuCV) has powder can be used successfully and advantage been established in the AuFs" ion. Salts ously in place of elemental niobium. It has prepared so far include Cs+AuF6", 02'|,AuE$" and also been established that pre-alloyed niobium- Xe2Fu*AuF6". The single crystal X-ray- zirconium powder can be prepared by hydriding structure of the latter has bef.n determined. arc melted buttons. This material is desir It appears that AUF5 is displaxd by SbFs able for future work on superconducting from its salts. materials. (N. Bartlett, p. 12) (M. R. Pickus, p. 147) Catalysis of hydrocarbon reactions has been shown to take place at an optimum rate on Experimental and theoretical techniques platinum surfaces that are characterized by previously used only to study congruent va the presence of ordered atomic steps and an porization reactions have been used to study an ordered carbon surface structure. incongruent decomposition reaction. The (G. A. Somorjai, p. 25) present studies on barium sulfate give results qualitatively similar to the previous study with calcite in that reaction rates are of the Memory filaments in As-Te-I amorphous semi order cf 15 of the thermodynamically allowed conductors have been identified as whisker rates and unaffected by a layer of oxide like crystals of tellurium. product up to 1 mm thick. The rate, however, A magnetron type sputtering apparatus has appears to be limited by a different process been constructed for the production of ZnO than was calcite decomposition. films for possible application as a low cost solar photovoltaic cell. (A. V. Searcy, p. 155) (J. Washburn, p. 97) A most significant achievement has been the General equations for the velocity of a clarification of the SiOj-AljOj phase diagram. dislocation gliding through a random array It has been shown that there are several of point obstacles at finite temperature have metastable diagrams which have not been pre been derived and the thermal "jerkiness" of viously proposed and that the original Bowen the motion explained. and Greig diagram is the stable one. A new interpretation of the surface stress (J. A. Pask, p. 160) The intrinsic nonstoichiometry has been Pair-quasiparticle potential differences established for the binary system ft>Zr03 - have been observed in superconductors. The PbTiC^. The techniques developed to determine results are in good agreement with a theory nonstoichiometry make it possible to fabricate involving the imbalance of electron-like and lead zirconate titanate ceramics with con hole-like excitations, a situation that has trolled defect structures.' not been previously discussed. The results (R. M. Fulrath, p. 169) are applicable to any superconductor which has been disturbed from equilibrium by a current A simple two-cavity dye laser difference involving a quasiparticle contribution. frequency generation system has been used to (J. Clarke, p. 197) generate tunable far infrared radiation over the frequency range from S to 190 cm"*. Since A tunable dye laser has been used to study the dye system can be operated in a repetitive resonance Raman scattering and luminescence in ly pulsed mode, this system provides a prac solids. Detailed information is obtained si tical tunable spectroscopic source. A total multaneously about the electronic band struc of five different types of far infrared reso ture, the phonon spectrum, and the electron- nant modes have been observed in alloys of the phonon interaction for a number of crystals. antiferromagnetic iron group fluorides The tuning range of the far-infrared output (Fe^h)Fj and (Co,Nh)F2. All of the resonant in optical mixing has been extended to a frequencies are shown to be in good agreement range from 0 to 200 cm"*. Information about with theory. This work answers a 10-year old the band structures of several semiconductors question of the source of extra modes some and metals has been obtained by analyzing the times seen in these classical materials. derivative spectra of the solids. (P. L. Richards, p. 188) Of. R. Shen, p. 205) -1-
A. INORGANIC CHEMISTRY
William L. Jolly^ Principal Investigator approximation that AE = 0 for the interchange of a pair of equally charged atomic cores 1. X-RAY PHOTOELECTRON SPECTROSCOPY between two different chemical species), the chemical shifts are estimated with an a. Chemical Shifts in Boron Is Binding average error of ±0.57 eV. (See Fig. 1.) Energies of agree Gaseous Compounds' There is no obvious correlation between the Is binding energy chemical shifts and published Patricia A. Finn and William L. Jolly llB NMR chemical shifts. Recently we reportedl the nitrogen Is electron binding energies for a series of gaseous nitrogen compounds and correlated the Abstracted from J. Am. Chem. Soc. 94_, 1540 experimental values with calculated atomic (1972). charges and thermodynamic data. In this study we have measured the boron Is electron 1. P. A. Finn, R. K. Pearson, J. M. binding energies of a series of gaseous boron Hollander, and W. L. Jolly, Inorg. Chem. in, compounds in order to provide data for 378 (1971). further tests of the correlation of binding energies with atomic charges and thermodynamic data. We have also compared the boron Is binding energies with boron-11 nuclear mag netic resonance chemical shifts. b. The Nitrogen Is Binding Energies of Transition Metal Nitrosyis* Chemical shifts in boron Is electron binding energies for gaseous BF3, BCI3, Patricia A. Finn and William L. Jolly
B[0CH3)3, B2H5, B(CH3)3, BH3CO, and H3BN(CH3)3 are shown to be linearly related to boron The bonding and electronic structure of atom charges estimated by Pauling, CNDO, and transition metal nitrosyis have been fre extended Huckel methods. The best correlation quently studied in recent years. Two limiting is found with the extended Huckel charges, situations have been identified: linearly with inclusion of the interatomic "Madelung coordinated NO* groups and angularly potential" (average deviation from straight- coordinated ("bent") NO" groups. We have line relation, ±0.83 eV). By appHcation investigated the X-ray photoelectron spectra of the thermodynamic method (based on the of a series of these compounds to attempt correlations of the nitrogen Is electron binding energy with structure, electronic 1 1 • I—I- —1—1—r- - features, and N-0 stretching frequency. 0- '•^ - The data are presented in Table I. It was found advantageous to categorize the compounds in terms of the me-al d-electron -2 - z - configurations by making the arbitrary *r assumption that the nitrosyl groups were >s N0+ ions. (Although this assumption is at _A _ illogical for the few compounds that contain - ' .3 - "bent" NO groups, we made the assumption to simplify the classification of the compounds.) 5 -6 - V /•4 - , 1 We draw the overall conclusion that there -8- - is a definite correlation between the 7 nitrogen Is binding energy of a nitrosyl V group and the electron density on that group. -10 - ,/ Bent nitrosyis have low binding energies. 1 • 1 1 . 1 1 1 1 Linear nitrosyis can have either low or high -11..0 J -8 -6 -4 -2 0 binding energies, depending on the extent Thermochemical Energy, eV of TT back-bonding from the metal atom. The binding energies are at least roughly Fig. 1. Plot of boron Is binding energies correlated with the N-0 stretching vs. thermodynamically estimated binding frequencies: when one is high or low, so energies. (XBL 716-6854) is the other. -2-
Table I. Nitrogen Is binding energies for some transition metal complexes.
Nitrogen No. of d Is binding Compound electrons energy, (eV)
MoCl2(NO]2Cdiars) 6 399.6 trans-[CrCl(NO)(diars)?]C104 5 400.7 K3[CrCNO)(CN)5] 5 400.7
[(C6H5)4P]3[Mn(N0)(CN)s]-3H20 6 399.7 trans-[RuCl(NO)(diars)2]Cl2 6 400.0 trans -[FeCl(NO)(diars)2](CIO4)2 6 402.9
Na2[Fe(NO)(CN)5]-2H20 6 403.3 [Tr-CsHsCr(NO)2]2 7 400.7
Rhl2&\'OHPCCH3)CC6Hs)2]2 8 400.3
RhCl2(NO)(PCC6H5)3]2 8 401.5 CoCl2(NO)(diphos) 8 400.7 trans- [FeCl(NO) (diars)2]C104 7 400.0
[Co(N01(NH3)5]Cl2 8 400.7 trans-[CoCl(NO)(diars)2]Cl 8 400.5
[Co(NO)Cdiars)2](C104)2 8 402.3 401.7, 399.6 CoCl2p»)[PCCH3)CC6Hs)2]2 8 CoCl2(NO)[P(n-C4H9)3l2 8 401.5, 399.7
[Fe(NOKdiars)2]CC104)2 7 401.2, 399.6
{RuClCNO)2[PCC6H5)3]2}PF6 8 402.6, 400.2 {RhCNO)2[PCC6H5)3]2>PF6 10 401.1
tIr(NO)2[PCC6H5)3]2)PF6 in 400.2 Co(N'0)fPCC6HsD3h 10 400.0
CoCNO)(CO)[P(C6H5)3]2 10 400.8
Rh(NO)[P(C6H5)3]3 10 400.8
Ir(N0)[PCCaH5)3]2 10 400.3
IrCNO)CCO)[P(C6H5)3]2 10 399.6
Ru3(NO)2CC0)l0 10 400.4 NiCl(NO)[P(C6H5)3]2 10 399.8
Ni(N3D(NO)[P(C6H5)3h 10 399.6
Abstracted from Inorg. Chem. 11, 893 (1972). The results indicate that the internal bonding of the azide ion is essentially unaffected by coordination to a transition c. Nitrogen Is Binding Energies of Some metal atom; that is, the azide ion appears Azide, Dinitrogen, and Nitride Complexes of to be linked to the metal atom by an ionic Transition Metals bond with little polarization. A dinitrogen complex can be considered to be a hybrid of Patricia A. Finn and William L. Jolly two major resonance forms, M - N+ = N and M = N* * N~. Because the nitrogen atom We have used X-ray photoelectron spectros directly attached to the metal atom has a copy to determine the nitrogen Is binding formal positive charge in both forms, we energies of azide, dinitrogen, and nitride expect decreased electron density on this complexes of several transition materials. nitrogen atom and a correspondingly higher Such data are of interest because the binding energy. The nitrogen Is spectrum binding energy shifts can be interpreted in of trans-fRu(N2)Cl-(diars)2J consists of terms of the interaction of the ligands with two peaks; we assign the higher binding the transition metal atoms. The data are energy to the nitrogen atom directly attached presented in Table I. to the ruthenium atom. Leigh et al., made -3-
Table I. Nitrogen Is binding energies for some transition metal complexes.
Nitrogen Is binding Intensity Separation Compound energy, (eV) ratio (eV)
NaN3 403.7, 399.3 1:2 4.4
Ni(N3)(NO)fPCC6HS)2]2 403.8, 399.6 1:2 4.2 trans-[Ru(N3)Cl(diars)2] 403.9, 399.2 1:2 4.7 trans- [Co (N3) 2 Cdiars) 2KIO4 403.2, 399.1 1:2 4.1
{Cu(N3)[P(C6H5)3]2)2 403.7, 399.2 1:2 4.S
[Rh(N3)(CO)2]2 403.2, 399.3 1:2 3.9
[AsCC6H5)4]2[Fe(N3)s] 402.8, 398.7 1:2 4.1
trans-[Ru(N2)Cl(diars)2] 402.3, 400.7 1:1 1.6 Re(N2)Cl(diphos)2 399.9, 397.9 1:1 2.0
[C6HS(N25]BF4 405.1, 403.8 1:1 1.3
K3[Ru2N(Cl)8(H20)2] 399.8
ReNCCl)2[P(C6H5)3]2 398.8
a similar assignment for [Re(N2)Cl(diphos)2]. The nitrogen Is binding energies of Abstracted from a paper in Electron K3IRU2N(C1)8(H 0) ] and ReN(Cl) [PCC6H5)3]2 Spectroscopy, edited by D. A. Shirley (North- are of interest2 becaus2 e these complexe2 s con Holland, Amsterdam, 1972), pp. 629-645. tain nitride ion ligands that are bridging e. The Correlation of Core Binding Energies and nonbridging, respectively. In the former by a Potential Model Based on "Half-Ionized"' complex, two metal atoms compete * for the Cores and a New Method for Calculating electron density of the bridging nitride; Atomic Charges" therefore a higher binding eneTgy is observed for this compound even though the complex Ifilliam L. Jolly, Winfield B. Perry, and has a 3- charge. Gregory R. Andersen If we make the approximation that the atoms Abstracted from Inorg. Chem. 1^, 1434 (1972). of molecules are non-overlapping conducting spheres, we can estimate core electron 1. G. J. Leigh, J. N. Murrell, W. Bremser binding energies from estimated atomic and W. G. Proctor, J. Chem. Soc. D, 1661 charges by using the relation (1970). 1 kQ • V + *, "B d. The Estimation of Core-ElectCore-Electro: n Binding- „ = t of the Energy Shiftta-^r.s m Using the Concept 1 where Q is the charge of the atom which loses Equivalence ot Equally Charged OCores " the core electron, V is the Coulomb potential energy of an electron at the hypothetical William L. Jolly vacated site of that atom in the midst of the other atoms of the molecule, and k and This review article summarizes the theory J are empirical constants. The energy V is and applications of the "equivalent core" calculated from the relation V - 2(q/r), in or "thermodynamic" method of estimating which q is the charge on an atom, r is its binding energy shifts. The predictive power distance from the site of the atom vacancy, of the method is demonstrated for gaseous and the sum is carried out over all the compounds of boron, carbon, nitrogen, oxygen, atoms. Implicit in the equation is the as and xenon, fOT gaseous monatomic ions of sumption that the core binding energy is a sulfur and chlorine, and for solid compounds function only of the initial state of the of boron, carbon, nitrogen, and iodine. The molecule. Hcwever, we know that during the use of the method for estimating chemical time required for the ejection of a core reaction energies from binding energy shifts electron from an atom in a nclecule, the is illustrated. valence electrons of the molecule shift toward the nucleus of the atom in which the 2. BORON HYDRIDE CHEMISTRY core hole is created. This concurrent valence electron relaxation process is believed to be a. A Kinetic Jtudy of the Intermediates in essentially complete in the time of the the Hydrolysis of the Hydroborate Ion* photoelectric process. Therefore, if we wish to calculate accurately the binding Francis T. Wang and William L. Jolly energy by a hypothetical "sudden" process in which the valence electrons are assumed to The hydroborate ion (otherwise Known as remain fixed, we cannot use the valence borohydride, tetrahydroborate, and electron distribution of either the initial tetrahydridoborate) undergoes hydrolysis molecule or of the final core-hole io . Use in aqueous solutions to give boric acid of the valence electron distribution of the below pH 9 and borate above pH 9. initial molecule would yield a binding energy too high by the "relaxation energy"; use of BH " + H+ + 3H 0 •* B(0H) + 4H the valence electron distribution of the 4 2 3 2 final ion would yield a binding energy too low because of the concentration of negative BH " + 4HjO •* B(0H) " + 4H . charge around the core. However, some 4 4 valence electron distribution between these two extremes, probably close to the average distribution, will give the right energy. A variety of experimental data have shown We have developed a procedure for correlating that the hydrolysis proceeds in four steps, and predicting core electron binding energies with the intermediate formation of trihydro-, based on this data. dihydro-, and monohydroboron species. In the present study we have shown that the Using the above theory as a basis, we have kinetics of the four consecutive steps of devised a simple empirical method, involving the acid hydrolysis of hydroborate can be electronegativity parameters, for estimating separately studied by appropriate adjustments atomic charges. The method permits us to of the reaction temperatures and the hydrogen correlate 129 different binding energies ion concentrations. We have prepared acidic (for compounds of boron, carbon, nitrogen, solutions in which the m.nohydroboron and fluorine) with an overall standard intermediate was the only boron-hydrogen deviation of ±0.61 eV. species present and solutions in which either the dihydro- or trihydroboron intermediate was the major boron-hydrogen species present. Abstracted from Faraday Disc. Qiera. Soc., Solutions containing the species BH(0H)3", 54 (1972) and LBL-1417. BH2(OH)2", or BH3OH- were prepared by the addition of excess sodium hydroxide to acidic solutions containing the appropriate f. The Use of X-Ray Photoelectron Spectros intermediates. The kinetics of the copy to Assess dOribital~Participation in hydrolysis of these anions was studied. Coimxnmds of Silicon and Germanium* The stepwise hydrolysis of hydroborate Winfield B. Perry and William L. Jolly has been studied in cold 88:12 vol % methanol- water solutions in the hydrogen ion concen Chemical shifts in core binding energies tration range 0.1-1.1 M. At -78°, BH4- have been determined for analogous gaseous rapidly hydrolyzes to E2OWZ, "hich, in turn, compounds of carbon", silicon, and germanium. hydrolyzes to BH2-(H20)2 according to the These data, in conduction with the atomic rate equation -d In [H20BH3]/dt = 0.0015 sec-1 charge potential model and both CNDO/2 and + 0.0016[H+] sec"1 M"1. The rate data for electronegativity parameter charges, indicate the hydrolysis of BH2(H2°)2+ solutions at that the silicon and germanium atoms in these -36° are consistent with the following rapid compounds, except for SiEfy and Gefy, have equilibrium (K = 6.4): H* + H2OBH2OH * unexpectedly negative charges. The results BH2(H20)2+- The BH2(H20)2 ion is stable are consistent with the existence of toward hydrolysis, whereas its conjugate significant p7r-d7r bonding in these compounds. base hydrolyzes to H20BH(0H)2 according to the rate equation -d In [H2OBH20H]dt = 0.017 sec"1. At -36°, H20BH(OH)2 hydrolyzes n to B(0H)3 according to the rate equation Abstracted from Chem. Phys. Letters, -d In [H2OBH(0H)2]/dt = 3.3 * 10"4 sec"1.
17, 611 (1972). The species H2OBH3, BH2(H20)2 , and -s-
H20BH(OH)2 are converted to the anions asymmetric cleavage of diborane by water and
BH3OH", BH2(OH)2~, and BH(0H)3", respectively, alcohol. by the addition of hydroxide. These anions undergo hydrolysis in nonbuffered, strongly alkaline solutions according to the rate Abstracted from Inorg. Chem. 11, 1941 (1972). equations -d In [BH30H"]/dt = 1.8 x 1Q"4 sec"1 at 20°; -d ln[BH2(0H)2~]/dt4.2 * ig-5 see"! 1. R. E. Davis and J. A. Gottbrath, Chan. 3 Ind. (London), 1961 (1961); R. E. Davis and and -d In [BH(0H)3"]/dt = 1,1 « 10" sec-1 at 0°. The boron-11 NMR spectrum of BH3OH- J. A. Gottbrath, Inorg. Chem. £, 1512 (1965). is a 1:3:3:1 quartet, with JB-H = 87 Hz 2. W. L. Jolly and T. Schmitt, J. Am. Chem. centered 12 ppm upfield from the borate Soc. 88, 4282 (1966). singlet.
c. Preparation of the BH3OH" Ion with a "Quenching Flow Keactor"~ Abstracted from Inorg. Chem. U, 1933 (1972). Janice B. Reed and William L. Jolly b. The Asymmetric Cleavage of Diborane by Water. The Structure of Diborane" Dihydrate* The work of Wang and Jolly1 indicated that, in the room-temperature reaction of H* and Patricia A. Finn and William L. Jolly BH4", the intermediate H2OBH3 would be expected to have a lifetime of the order of The cleavage of the bridging hydrogen magnitude of milliseconds. Therefore an bonds in diborane by a Lewis base, X, can apparatus was set up for mixing streams of be either symmetric, to give BH3X, or H* and BH4" solutions and then quickly asymmetric, to give BH2X2+ and BH4". In the mixing the resulting reaction mixture with reaction of ammonia with diborane, both a stream of OH" solution. With this apparatus types of cleavage have been observed. With we have prepared BH3OH" solutions, free of increasing methyl substitution in the other borohydride species, at concentrations methylamine series, there is an increasing as high as 0.21 M in 21% yield and 0.052 M tendency for symmetric cleavage. Various in 52% yield. Kinetic data for the hydroTysis ethers cleave diborane symmetrically, whereas of the EH3OH" at 20° yield a half-life of dimethyl sulfoxide cleaves diborane 4.1 hours. We plan to investigate the asymmetrically. It appears that it is not chemical properties of the BH3OH" species. yet possible to predict the course of such With this technique it should be possible to reactions with certainty. prepare, by use of slower quenching, the other borohydride intermediates, BH2(0H)2~ In studies of the reaction of diborane andBH(0H)3". with water and alcohols at room temperature, no intermediate cleavage products have been The general utility of the quenching isolated because complete hydrolysis occurs method has been demonstrated2 by preparing rapidly. Davis and Gottbrath* were able to the peroxynitrite ion (CONCT) in '1% yield isolate sodium hydroborate from the reaction from H2O2 and NO2". In this synthesis, the of diborane with aqueous sodium hydroxide at very unstable HDONO is prepared from H2O2 and 0°, and Jolly and Schmitt2 prepared potassium HN02, and it is then quenched with OH" which hydroborate by the reaction of diborane with converts it to the relatively stable CONO". powdered potassium hydroxide at -30°. These Further development of this synthesis is results are consistent with asymmetric planned. cleavage of diborane and prompted a further study of this type of reaction. This paper describes the preparation and characteriza tion of the initial diborane-water reaction 1. F. T. Wang and YJ. L. Jolly, Inorg. Chem. product. 11, 1933 (1972). 2. Work carried out by Mr. H. H. Ho in The dihydrate and diethanolate of diborane Chemistry 106, Fall, 1972. have been prepared at -130° and their compositions have been established by stoichiometric measurements. The dihydrate 3. SULFUR-NITROGEN CHEMISTRY is assigned the structure BH2(H20)2+BH4" on the basis of low-temperature infrared a. The Reaction of Heptasulfur Imide and spectra of the deuterated and nondeuterated Diborane* species. A study of the evolution of hydrogen from the reaction of diborane with Marshall H. Mendelsohn and William L. Jolly a solution of HC1 in a water-methanol mixture at -78° has provided further evidence for the It is known that heptasulfur imide reacts •6-
with boron trichloride and boron tribr&mide S7NH in tetrahydrofursm at 0°, the reaction to give S7NBCI2 and S7NBEr2, respectively. does not stop with the formation of S7NBH2, In this note we report the results of our as evidenced by the continued slow evolution study of the reaction of heptasulfur imide of hydrogen. Presumably the following subse with di'jorane. This study was undertaken quent reaction occurs: to determine whether the reaction gives an S NH + S NBH - (S N) BH • Hj analogous product (S7NBH2), a borane adduct ? 7 2 7 2 (S7NH • BH3), or a diborane derivative such However, the reaction was incomplete after as S7NB2H5. one month and was not further studied.
The reaction of heptasulfur imide with diborane was carried out in various ether Abstracted from Inorg. Chem. 11_, 1944 (1972). solvents, at either room temperature or 0°. The experimental results show that, with b. The Structures of Sulfur-Nitrogen diborane in excess, 0.5 mol of diborane Compounds* reacts per mole of S7NH and that 1 mol of hydrogen is formed per mole of S7NH. This William L. Jolly stoichiometry corresponds to the reaction
S7NH + 0.5B2H& » S7NBH2 + Hj. In this paper the structures of compounds which contain rings or chains of alternating When the reaction is carried out in sulfur and nitrogen atoms are discussed. diethyl ether at 0", some of the product During the last 15 years or so this class precipitates as a white, crystalline ether of compounds has frustrated and mystified adduct, S7NBH2 • Et2<5. Similar adducts can chemists because various structure determina be obtained from the other ethers by vacuum tions have often showed that structures which evaporation of the solvent and removal of had been predicted for these compounds were excess diborane. The compounds undergo wrong. However, enough structures of such rapid hydrolysis with excess water at room compounds are now known that it is possible temperature: S7NBH2 • ether + 3H20 -»• to systematize and rationalize the data.
S7NH + B(CH)3 + 2H2 + ether. The structures and bonding of the following The 11B NMR spectrum of S7NBH2 in diethyl species have been studied: S2N2, S2N2+, ether solution consists of a 1:2:1 triplet S2N22+, S3N3CI3, S3N3O3CI3, STNH, S6(NH)2, due to coupling with the two protons of the S5(NH)3, S4N4H4, S4N4F4, S4N42-, S4N4, 24 + BH2 group. The Raman spectrum of the diethyl S4N4 , S5N5*, S3N2O2, S3N2C1 , S3N3NPPh3, ether adduct has lines at 2419 and 2409 an"'-, S4N2, and S4N3*. These species are characteristic of terminal BH2 groups, and characterized by rings or chains of alterna lines at 766 and 750 cm"1, typical S-N ting sulfur and nitrogen atoms. The struc stretching frequencies. tures are rationalized by comparison with isoelectronic compounds of known structure and The only products that were stable (as by assuming delocalized pir bonding whenever determined by hydrogen evolution) were those possible. obtained from diethyl ether and tetrahydrofuran. The diethyl ether adduct is stable at room temperature in the absence of air and moisture, and the product from Abstracted from Advances in Chemistry tetrahydrofuran is stable at 0°. Hydrogen, Series 110_, 92 (1972). hydrogen sulfide, and ether were identified as decomposition products of the tetrahydrofuran and monoglyme adducts at 4. THE PREDICTION OF GAS-PHUSE LEWIS ACID- room temperature. BASE DISSOCIATION HEATS WITH AN EMPIRICAL FOUR-PARAMETER EQUATION* A white solid pyridine adduct, S7NBH2 C5H5N, w
log ^ . ^ • 6H, William L. Jolly a. X-Ray Photoelectron Spectroscopy
where K is the equilibrium constant for the We shall try to devise a method for obtain reaction of a base with a particular acid and ing the spectra of high temperature gaseous KQ is the constant for the corresponding species. The main problem to be overcome is reaction of a reference base. The parameters the effusion of vapor through the electron a and B are empirical constants characteris slit, and consequent fouling of the vacuum
tic of the acid, and En and H are independent chamber. The use of a beam or jet of vapor parameters for the base. More recently, molecules seems indicated. Drago and Wayland* used a similar equation to reproduce the heats of dissociation of We plan to investigate the binding energy Lewis acid-base adducts: shifts for compounds of nickel and iron in various oxidation states in order to determine the valence electron distribution in these compounds. m ' EAEB + CAC1 We are developing a new method for cal culating molecular electron distributions, Two empirical parameters, EA and CA, were based on electronegativity equalization. assigned to each acid, and two empirical parameters, EB and Cfi, were assigned to each b. Boron Hydride Chemistry base. The EAEB term was interpreted as a measure of electrostatic interaction energy, We shall e:\ploit the quenching flow and the CACB term was interpreted as a reactor as far as possible. Good syntheses measure of covalent bond energy. By of BH2(OH)2" and BH(0H)3" should be possible, application of the equation to reactions of as well as non-boron-containing species such neutral acids and bases in the gas state or as OONO". in poorly solvating solvents (for which the dissociation heats ranged from 0.5 to c. Sulfur-Nitrogen Chemistry 18.0 kcal/mol), it was possible to correlate the dissociation heats with an average error We hope to study quantitatively the of approximately ±0.1 kcal/mol. isomerization of the S7N" ion in ether solu tion (see Section 4-a) by means of UV and Dissociation heats are available for an visible spectrophotometry. An apparatus has extensive set of gaseous compounds, not been built for preparing the solutions and considered by Drago and lVayland, which may be transferring them to an absorption cell in considered as Lewis acid-base adducts. I*>st the complete absence of air. We plan to of these compounds are adducts of cationic study the spectrum as a function of concen acid's and anionic bases. tration and temperature. By use of the empirically evaluated parameters for 14 acids and 12 bases, the heats of 105 different dissociations are reproduced with an average error of 6. 1972 PUBLICATIONS AND REPORTS ±3.9 kcal/mol. The method gives poor results for adducts in which the dissociated bonds William L. Jolly and Associates are either weakened by lone-pair repulsions or affected by unusual ir-bonding effects. Journals and books 1. William L. Jolly, The Structures of Sulfur-Nitrogen Compounds, in Advances in n Chemistry Series 110, 92 (1972). Abstracted from Inorg. Chem. 11, 869 (1972). 1. R. S. Mulliken, J. Am. Chem. Soc. 74_, 2. William L. Jolly, The Estimation of Core- 811 (1952). Electron Binding-Energy Shifts Using the 2. J. 0. Edwards, ibid. 76_, 1540 (1954); Concept of the Equivalence of Equally Charged 78, 1819 (1956). Cores, in Electron Spectroscopy, edited by 3. R. S. Drajo and B. B. Wayland, ibid, 87., D. A. Shirley (North-Holland, Amsterdam, 3571 (1965). 1972), pp. 629-645. -8-
3. Patricia A. Finn and William L. Jolly, 2. William L. Jolly, X-Ray Photoelectron Chemical Shifts in Boron Is Binding Energies Spectroscopy and Intermediates in the of Some Gaseous Compounds, J. Am. Chem. Soc. Hydrolysis of Borohydride and Diborane, 94,1540 (1972). A. C. S. Visiting Scientist Program, California State Polytechnic College, Pomona, 4. William L. Jolly, John D. Illige, and Calif., Feb. 16-17, 1972, and University of Marshall H. Mendelsohn, The Prediction of Nevada, Reno, March 23-24, 1972. Gas-Phase Lewis Acid-Base Dissociation Heats with an Empirical Four-Parameter Equation, 3. William L. Jolly, Intermediates in the Inorg. Chem. 11_, 869 (1972). Hydrolysis of Borohydride and Diborane, Worcester Polytechnic Institute, Worcester, 5. Patricia A. Finn and William L. Jolly, Mass., April 12, 1972. W. L. Jolly, X-Ray The Nitrogen Is Binding Energies of Transition Photoelectron Spectroscopy, U.C.L.A., Los Metal Nitrosyls, Inorg. Chem. U, 893 (1972). Angeles, California, May 2, 1972.
6. William L. Jolly, Encounters in Experi 4. W. L. Jolly and K. A. Strcai, Reactions mental Chemistry (Harcourt Brace Jovanovich, of Hydrogen in Potassium Amide-Liquid New York, 1972). Ammonia Solutions, 3rd International Conference on Non-Aqueous Solvents, Michigan 7. Patricia A. Finn and William L. Jolly, State University, East Lansing, Michigan, Nitrogen Is Binding Energies of Some Azide, July 5-7, 1972. Dinitrogen, and Nitride Complexes of Transi tion Metals, Inorg. Chem. 11, 1434 (1972). 5. W. L. Jolly, Inorganic Applications of Photoelectron Spectroscopy, Gordon Research 8. Francis T. Wang and William L. Jolly, Conference on High Temperature Chemistry, A Kinetic Study of the Intermediates in the Andover, N. H., August 7-11, 1972. Hydrolysis of the Hydroborate Ion, Inorg. Chem. 11_, 1933 (1972). 6. W. L. Jolly, X-Ray Photoelectron Spectroscopy, University of Glasgow, Glasgow, 9. Patricia A. Finn and William L. Jolly, Scotland, September 5, 1972, and University The Asymmetric Cleavage of Diborane by Water. of Edinburgh, Edinburgh, Scotland, September The Structure of Diborane Dihydrate, Inorg. 7. 1972. Chem. 11_, 1941 (1972). 7. W. L. Jolly, The Correlation of Core 10. Marshall H. Mendelsohn and William L. Binding Energies by a Potential Model Based Jolly, The Reaction of Heptasulfur Imide on "Half-Ionized" Cores and a New Method for with Diborane, Inorg. Chem. 11, 1944 (1972). Calculating Atomic Charges, Faraday Discussion of the Chemical Society, Brighton, England, 11. William L. Jolly, ed., Metal-Ammonia September 12-14, 1972. Solutions, in Benchmark Papers in Inorganic Chemistry (Dowden, Hutchinson and Koss, 8. W. L. Jolly, X-Ray Photoelectron Stroudsburg, Penn., 1972). Spectroscopy, Frontiers in Chemistry Lecture, Wayne State University, Detroit, Mich., 12. William L. Jolly, The Correlation of October 30, 1972. Core Binding Energies by a Potential Model Based on "Half-Ionized" Cores and a New 9. W. L. Jolly, The Estimation of Atomic Method for Calculating Atomic Charges, Charges in Molecules, University of Windsor, Faraday Disc. Chem. Soc. 54, (1972). Windsor, Ontario, Canada, October 31, 1972.
13. Winfield B. Perry and William L. Jolly, 10. W. L. Jolly, X-Ray Photoelectron The Use of X-Ray Photoelectron Spectroscopy Spectroscopy, The Estimation of Atomic to Assess d Orbital Participation in Compounds Charges in Molecules, Intermediates in the of Silicon and Germanium, Chem. Phys. Letters Hydrolysis of Borohydride and Diborane, and 17, 611 (1972). A Comparison of Analogous Hydride Derivatives of Germanium and Carbon, presented during Visiting Professorship at the University of Western Ontario, London, Canada, November Papers presented 7-16, 1972.
1. W. L. Jolly, Thermochemical Estimates of 11. W. L. Jolly, Chemical Shifts in Core Core-Level Binding Energy Shifts, Symposium Electron Binding Energies, Symposium on on Electron Spectroscopy, Case Western Review Photoelectron Spectroscopy, Yale University, University, Cleveland, Ohio, January 10-11, New Haven, Conn., November 30, 1972. 1972. -9-
LBL report 2. E. R. Staricco, C. Riddle, and William L. Jolly, The Reaction of Potassium Germyl 1. Marshall H. Mendelsohn and William L. with Boron Trimethyl, submitted to J. Inorg. Jolly, Reactions of the Heptasulfur Imide Nucl. Chem. (LBL-826, April 1972). Anion, submitted to J. Inorg. Nucl. Chem. (LBL-820, August 1972). 3. William L. Jolly, W. B. Perry, and G. Anderson, A Method for Estimating Atomic Charges Based on Electronegativity Parameters, LBL-1417, December 1972.
S0$?75 -10-
Robert E. Conniak, Principal Investigator models are being undertaken. So far only two-dimensional cases with 72 particles have 1. OXYGEN-17 »K STUDY OF THE RATE OF WATER been attempted. Single-step and three-step EXCHANGE FROM PARTIALLY COMPLEXED NICKEL ION square wells have been calculated for a single radius. An 18-step well will be tried Thomas V. Rowland next and the variation of rate and reaction path with depth of well and radius of "metal In a continuing study of the exchange ion" will be studied before going to three lifetimes of water molecules in the first dimensions. It is hoped that the calculations coordination spheres of metal ions, partially will give information on the mechanism of complexed nickel ion is being investigated. exchange and the effect of radius and charge The 1:1 complex with iminodiacetate ion has of the "metal ion." The work is being done three positions occupied by water molecules, in collaboration with Dr. Berni Alder of the two of which are equivalent. Measurements Livermore Laboratory. of the line width of the oxygen-17 bulk water resonance are being made as a function of temperature to determine the effect of com- plexing on the water exchange rate and to see if the two kinds of coordinated water molecules exchange at appreciably different rates. The measurements have also revealed an unexpected line broadening near 0°C which 4. RAMAN SPECTRUM OF TRIIODIDE ION is being investigated. Kenneth Suslick
It is reported in the literature [Maki and 2. 170-NMR STUDIES OF WATER ^DLECULES IN THE Forneris, Spectrochim. Acta 23A, 867 (1967)] 3+ that triiodide ion in aqueous solution shows SECOND COORDINATION SPHERE OF Cr two Raman lines, which would require an unsym- metrical structure. With the availability of William L. Earl lasar Raman apparatus such measurements can be made much more precisely. The present Olson, Kawagawa, and Taube [J. Chem. Phys. study showed that in IT" solutions containing SI, 289 (1969)] have used the line broadening excess I there are indeed two lines, but oF 170 of bulk water in solutions of Cr(III) the intensity of one of these is a function to set a limit on the lifetime of water of the I" concentration, disappearing at low molecules in the second coordination sphere iodide. Therefore, the Raman data do not of Cr(III). The present work is directed give evidence of an unsyrametrical structure towards fixing the value of the lifetime by for triiodide ion. The line which varied combining both transverse and longitudinal with iodide concentration probably arises relaxation times for the 170. A preliminary from the presence of appreciable l^- jj, tne value of 3 x 10-10 f the lifetime at sec or solutions. 25° is being refined. These results will give the first measurement of the exchange lifetime for a second coordination sphere water molecule, a result of which is of interest in the understanding of mechanisms and rates in aqueous solutions.
5. RESEARCH PLANS FOR CALENDAR YEAR 1973
3. COMPUTER SIMULATION OF SOLVENT EXCHANGE Robert E. Connick FROM THE FIRST COORDINATION SPHERES OF METAL IONS The studies of water exchange rates for the first coordination sphere of metal ions will Robert E. Connick be continued, and it is hoped that several of the simple hydrated metal ions whose rates In order to understand better the complex are very fast can be measured. Water ex movements of the many particles involved in change on additional partially complexed ions the exchange of a first coordination sphere will be investigated. The computer modeling water molecule on a metal ion with thp -jlk of the exchange process will be extended to water, computer calculations of simplified three dimensions. -11-
6. 1972 PUBLICATIONS AND REPORTS LBL Reports Robert E. Connick and Associates 1. E. E. Genser and R. E. Connick, The Ex change of Iodide Ion with Triiodide ion Journal Studied by Nuclear Magnetic Resonance, LBL-669, July 1972. 1. James W. Neely and Robert E. Connick, Oxygen-17 Nuclear Magnetic Resonance Studies of Aqueous Nickel Ion, J. Am. Chem. Soc. 94_, 3419 (1972). -12-
Neil Bartlett, Principal Investigator 1. NOBLE-GAS AND HALOGEN CHEMISTRY a. Krypton Chemistry Douglas E. McKee and Neil Bartlett The more effective synthesis of KrF2 via the 10 MsV proton bombardment of Kr/F2 mixtures Fig. 1(b). Stereogram showing the arrange has made for a significant advance of KrF? + ment of the XeF3 Sb2F ' structural units chemistry this year. The KrF*" salt KrF^SE^Fii" u has been used to oxidize XeOF to XeOF5+: within the unit cell--view along b. 4 (XMT729-6983)
+ KrF* + XeOF4 - Kr + XeOFj ^
Abstracted from D. E. McKee, C. J. Adams, and Neil Bartlett, Chem. Coram., January issue 1973. b. Xenon Chemistry Christopher J. Adams, Donald D. Gibler, Kevin Leary, Douglas E. McKee, Boris Zemva, Thomas J. Richardson and Neil Bartlett (i) Xenon(H) Chemistry (Zemva): Inter action of XeF2 with PdF* yields PdF4 complexes of which: 3XeF2-PdF4, XeF2-PaF4, XeF,-2PdF4 are certain and 4XeF2-PdF4 and 2XeF2-PdF4 possible. (T. Richaidson): The hydrolysis of + Xe2F3 RuF6" yields almost a 100% yield of RuCty, suggesting that the hydrolysis of Xe(II) fluoro-species may be an effective way to
Fig. 1(c). The XeF, ion and its close contact with the Sb,Fii" ion. ' (XBL 729-6918)
prepare higher oxides. Hydrolysis of Xe2F3+IrF(j" produces some volatile iridium
species but so far Ir04 has not been con firmed.
+ (ii) Xenon(IV) Compounds*: The XeF3 salts XeF3+SbF6" and Xe&^Sb^FvT have been prepared (McKee) and characterized by Raman spectroscopy (Adams). A full crystal structure determina + tion (McKee with Zalkin) of the XeF3 Sb2F11" has been carried out* and is shown as Fig. 1.
+ Fig. 1(a). The structural unit XeF, Sb2Fi1". (XBL 729-6975) -13- and X^Fji^'AuFg" have been prepared and characterized by Raman spectroscopy (Leary). F»-A*-FI' i».ip)* n-w-rz ».«*)• F4-I(2-FI 7M(4)' The crystal structures of the first and the Ft ».
The preparation of Xe2Fn AuFg" (Leary) is especially noteworthy since it represents the first Au(V) compound. The crystal structure
Fig. 3. The structural unit of Xe2Fn AuF6" Cestimated standard deviations in parentheses). (XBL 7210-7119)
+ 2 Fig. 2(a). The formula unit in [XeFs ]2PdF6 (XBL 724-6188) -14-
of the compound not only structurally defines The Raman spectra of the AuF6~ and PdFg2" the AuF6" ion for the first time but also salts of Xe2Fn+ and XeFr+ given in Table I establishes the geometry of the Xe2Fu+ ion. show the clear relationship of Xe^FTi* to + The structure is sham in Fig. 3. XeF5*. The XeFc has been further defined vibrationally (Adams). Comparison with data The cation Xe2Fii* is very like two XeFj+ for isoelectronic molecules reveals unusual groups joined by an F". The short Xe-F vibrational properties for XeF5+, in that the bridging distance of approximately 2.2 A symmetric stretching fundamentals lie lower indicates, however, that there is some cova- than in IF5, though the bond lengths of XeFj* lency in the interaction. The complex cation are the shorter. The probable reason lies is probably best described as a resonance with ligand-ligand repulsions in the crowded + + + hybrid of (XeFs F-XeFs ), (XeF$XeF5 ), and XeFs* molecule, and this should be evident in (XeF5+XeFg) canonical forms, with the first the values for interaction constants given by form dominating. a full normal coordinate analysis. A similar
TABLE I Raman Spectra of X^ll* salts and Related Species (shifts in cm" )
C JE5 i2 (B&jEFst (XeF5!)2HF6t tezru%M6t SMz^JMi. csW6:
•668 w 710(v,) vs . 653 s .... L651 vs 631(^1 sh .
r616 vw "615 w 614^) vs . . . . [626 w • * 1-606 w .606 w 1.600 s
•591 s 595 vs 595^) vs
602 573^) s . . . . 558 ms . . . . 568 s 564(vg) ms . . . 535 S .... 546 w 530(2^) VW (-425 vw T412 vw 370(vo) w .... 400 w ~* L396 vw L396 vw 375 w 356 w 318(v3) m 309 w .... 296 w (b) 290 w 274^) w 269 w ... . 270 w (b) 243(^1 ms . . . 245 w . . . . 245 w (b) 221 m , 224^) s (a) L. E. Alexander and I. R. Beattie, J. Chem. Soc.(A). 1971, 3091. -15- Table II Raman and Infrared Spectra of the Tetrafluoroiodate[III] a XeF C 5 Assignment in D,h symmetry Bu"4NIF4 4 Solid Solu. in MeCN gas R r lr R lr 511 ,s 519,s,p 554.3 -(a. J MF. in-phase str. 291 Z^ip >1F4 out"of"Plane ben 47Ss 586 -(e) MF4 degenerate str. (a) P. Tsao, C. C. Cobb, and H. H. Claassen, J. Chem. Ehys. 5£, 5247 (1971). effect has been observed for the series SbFg", combine together in the vapor phase to give + TeFg, IF6 . These and related studies promise a white solid, composition N0F-IF7, with a to throw light on the problem of structure and vapor pressure of — 300 mm at room temperature, bonding in XeFg itself. the Raman spectrum of the solid at ca. -80°C shows a line at 2240 cm"1 attributable to N0+, The synthesis of XeoFji+AuFg" points to and a characteristic pattern of lines in the properties of XeF6 which nave Sot been ex 1-F stretching region. Reaction of this com ploited previously. The hexafluori.de is not pound with cesium fluoride at -7S°C in excess oxidizable and melts at the relatively low NOF as solvent produces a cesium fluoro- temperature of 49.5°C. It is also a moderate iodate(VII), contaminated with CsNOj and ly good base. It is, therefore, valuable as residual CsF. The Raman spectrum of CsIFg in solvent and base in promoting unusually high the I-F stretching region closely resembles oxidation states of the elements (see NOF-IFy. On the basis of the spectroscopic Research Plans for 1973). evidence the solid is believed to be pre dominantly CsIFg, and efforts will be made to + The XeOF3 salts^XeOFj+SbFg" and XeOFj+Sb^j" obtain its crystal structure. have been prepared (McKee), and Raman studies (McKee and Adams) have demonstrated the salt (ii) IF4* (Gibler), has been shown to formulations. C?v symmetry in the salt IF.'*'SbFfi* which is orthorhombic [a - 8.29(1), B = 8:34(1), K. Leary and N. Bartlett, J. Chem. Soc. Chem. c * 10.34(1)] and not tetragonal as had been Conm** . 903 (1972). thought previously. D. E. McKee, C. J. Adams, and N. Bartlett, J. Chem. Soc. Chem. Comm. 26 (1973). 2. FLU3RIDES, OXYTLU0RIDES, AND OXIDES OF THE TRANSITION METALS c. Halogen Chemistry Alain R. Tressaud, Kevin Leary, and (i) Iodine Fluoranions (Adams): Neil Bartlett 11 IF," - Bu . NIF4 was prepared by the reduction a. Fluorides of Rhodium and Iridium (Tressaud) of IFc with Bun4 NI. Infrared and Raman data shoved the anion to be square planar, and thus No noble-metal tetrafluoride crystal struc isostructural with XeF. (Table II). ture is presently known. The tetrafluoride of IFg" - IF7 is a much weaker Lewis acid than iridium has been established and magnetic XeFj; direct reaction with alkali metal fluo studies (from 4°K to room temperature) have rides ii hampered by its poor solvent proper confirmed the low spin d5 system consistent ties. Attempts to prepare octafluoroiodates (VII) with iridium(IV). Some subtle antiferro- centered on the reaction of NOF and IF7, which magnetic-like coupling shows up below 90°K. -16- Attenpts to grow single crystals of IrF4 ligands. The F atom packing in the (^5)4 by recrystallization fron IrFr have failed, structure is essentially hexagonal-close- but good quality pcwder data support the rela packing. The Rh-F-Rh bridge angles of ~ 136° tionship to the ThCl4 unit cell. The body- indicate considerable covalency in the bonding centered pseudo monoclinic unit cell (true and represent the greatest possible departure unit cell F orthorhombic) has a - 6.680, from linearity allowable, without gross dis b - 5.672, c - 6.680 (all ± 0.003 A), tortion of the RhFg units. 6 - 92.67*. Since IrF/t, PtF4, and PdF^ are isomorphous they may well be isostructural and since the 3. RESEARCH PLANS FOR CALENDAR YEAR 1973 powder data indicate a similarity to the ThCl4 (UCI4) data there is the possibility of Neil Bartlett that type of structure occurring in these fluorides. Since the hexafluoride, penta- Hoble-Gas and Related Chemistry. RrF* fluoride, and trifluoride of iridium all show promises to be an oxidizer of extraordinary octahedral coordination of the metal atom, power. The possibility exis'.s of C1F,*, the adoption of the ThCI.4 structure (where Th BrFg+ syntheses and transition element is light-coordinated in chlorine) is un oxidations to previously unknown oxidation expected. states. Rhodium tetrafluoride is not isomorphous Efforts will be made to isolate single with the form of M4 which is stable at room crystals of an XeCtfv* salt so that the cation temperature; nor is HhF4 isomorphous with RUF4. can be property characterized. The related base XeOFg will be sought. Suitable single crystals of RI1F5 have been obtained for X-ray structure analysis and a Raman studies of XeF6 in a variety of high-precision structure has been obtained. solvents (HF, WFg, etc.) will be undertaken (Fig. 1) This establishes that RhFs is in order to better define the relationship isostructural with RuFs and C5F5 but, for + of molecular XeF6 to !fe2Fij and XeFs*. the first time, the precision permits mean XeF(j will be further exploited as a solvent ingful discussion of the bonding. As may be and base (see belowl seen from Fig. 1, the Eh~F-Rh bridge bonds (~2,0 A) are significantly longer than the non- The interaction of XeF2 with XeF,+ and bridging (1.8 A). This is consistent with single-electron Rh-F (bridge) bonds and XJ3F5 will be further investigated with electron-pair bonds for the uniquely bound F regard to the F" donor properties (and investigated with regard to the F" donor properties (and bonding) in the xenon fluor ides. Efforts will be made to isolate salts of Xe3F42+. Synthesis of new Xe(II) compounds of formulae FXeR, or XeR2 (where R = SF5, OSF5, etc.) will be attempted. To this end the synthesis and properties of SFsBr will be explored. The interaction of XeF, and its salts with water will be exploited as an oxygen atom source, and species such as IrC"4, IrOF5, °s02F4, OB^g will be sought. The crystal structure of CsIFs" (or a relative) will be attempted. Transition. Lanthanide, and Actinide Fluorides and OxyfluoridesI Xenon hexa fluoride- tluorine mixtures will be used to attempt the preparation of Np(VII), Pu(VTII), and Pr(V) salts. The isolation of AuFc will be undertaken and the synthesis of AuFg will be attempted. The synthesis of AuFg would provide a more Fig. 1. The EhJv tetramer. (XBL 728-6758) powerful oxidizer than PtFg. AuFg may oxidize -17- Kr spontaneously at ordinary temperatures and Papers presented pressures. It is also conceivable that it could stabilize the ArF* ion in the salt 1. N. Bartlett, "Bonding in Noble-Gas Chemis ArF* AuFg". try," Simon Fraser Universityt March 28, 1972; Oregon State University, March 29, 1972; Hopefully the XeF2 hydrolysis studies will University of California, San Diego, June 13, provide a convenient synthetic route to higher 1972; University of Southern California, oxide and oxyfluorides, but more traditional December 1, 1972. oxyfluorinations will also be undertaken in an effort to prepare NpOFs etc. 2. N. Bartlett, European speaking tour; "The Highest Attainable Oxidation States of the Single crystal and Raman studies of OSO3F2 Noble-Metals," University of Bordeaux, June will be undertaken to establish that the con- 26, 1972; "Bonding in Noble-Gas Compounds," pound is a fluorine bridged polymer. Similar Jozef Stefan Institute, Ljubljana, Yugoslavia, studies will be carried out with ReG^Fj. June 27, 1972; "Bonding in Noble-Gas Compounds," Lehrstuhl B fur Anorganische Chemie der PdF4 and its xenon fluoride adducts will be Technische Universitat, Braunschweig, Germany, further studied in an effort to prepare July 1, 1972; "The Highest Attainable Oxida ++ Xe3F4 salts and XeF5*.XeF2 adducts and to tion States of the Elements," Anorganisch- obtain a crystal structure of PdF,}. Chemisches Institut der University, Gottingen, Germany, July 5, 1972; "Fluoride Bridge Bonding" and '"Bonding in Noble-Gas Compoinds," Institut fur Anorganische md Analytische 4. 1972 PUBLICATIONS AND REPORTS Chemie, Justus Liebig-Uhiversitat, Giessen, Germany, July 7, 1972; "Fluoride Bridge Neil Bartlett and Associates Bonding," Universitat des Saarlandes, Saabrucken, Germany, July 10, 1972; "Fluoride Journals Bridge Bonding," Universitat Dortmund, Dortmund, Germany, July 12, 1972; Bonding in 1. N. Bartlett, M. Wechsberg, G. R. Jones, Noble Gas Compounds," Institut fur and R. D. Burbank, The Crystal Structure of Anorganische Chemie der Technischen Hochschule Aachen, Aachen, Germany, July 14, 1972. Xenon(II) Fluoride Fluorosuljfliate, FXeOS02F, Inorg. Chem. 11, 1124 (1972) (LBL-170). 3. N. Bartlett, "Noble-Gas and Related 2. K. Leary and N. Bartlett, A New Oxidation Chemistry," 14th Annual Stfjner Conference of State of Gold: The Preparation and Some Proper the California Association of Chemistry ties of [AuF0]- Salts, Chem. Coram. 903 (1972] Teachers, Asilomar, California, August 25, (LBL-870). 1972. 3. Neil Bartlett, Noble-Gas Compounds, 4. N. Bartlett, C. J. Adams, K. Leary, and Endeavour XXH [114], 107 (1972) (LBL-834). D. E. McKee, "Acid-Base Properties of &F4 and XeOFi," Annual ACS Meeting, New York, 4. D. D. Gibler, C. J. Adams, M. Fischer, August 29, 1972. A. Zalkin, and N. Bartlett, Structural Studies of Tri£Luorosulfur(IV)yl, [SF,]+, Salts In 5. N. Bartlett, "Concerning the Nature of tire cluding the Crystal Structure of [SF3]*[BF4]", Bonding in Noble-Gas and Related Compounds," Inorg. Chem. 11, 2325 (1972) (LBL-469). Plenary Lecture, Dalton Division Symposium, Autumn Meeting of the Chemical Society, 5. M. Wechsberg, P. A. Bulliner, F. 0. Sladky, Nottingham, England, September 26, 1972. R. Mews, and N. Bartlett, Fluorosulfates and Perchlorates of Xenon(II) and the Salt 6. N. Bartlett, "Concerning the Nature of + [(FXeO)2S(0)F] [AsF6]-, Inorg. Chem. 11, 3063 Bonding in Noble-Gas Contpovnds," ACPS (1972) (LBL-168). Pacific Regional Meeting, Jack Tarr Hotel, San Francisco, October 16, 1972. 6. N. Bartlett, Noble-Gas Chemistry and Its Significance, Jahrbuch der Akademie der LBL reports Wissenschafter in Gottingen, 1971, Vandenhoeck and Ruprecht, Gottingen, Germany (Dannie- 1. N. Bartlett, M. Gennis, D. D. Gibler, Heineman Prize Address), (Akademie-Jahrbuch) B. K. Morrell, and A. Zalkin, The Cryitjl (LBL-416). Structures of [XeF]*[RuF6]" and [XeFsJ*[RuF6]-, LBL-850, August 1972. -18- 2. B. K Morrell, A. Zalkin, A. Tressaud, 5. K. Leary, A. Zalkin, and Neil Bartlett, + _ and N. Bartlett, The Crystal Structure of The Crystal Structure of Xe7F11 AuF6 and + Rhodium Pentatluoride, LBL-1154, December 1972. the Raman Spectrum of Xe2F1i , LBL-1189, October 1972. 3 D. E. McKee, A. Zalldn, and N. Bartlett, + The Crystal Structure of DteF3] [Sb2Fn]-, 6. D. E. McKee, C. J. Adams, and N. Bartlett, LBL-1160, October 1972. The Preparation and Raman Spectra of the Salts + _ + + XeF3 SbFfi , XeF3 Sb2Fii-, XeOF3 SbF,-, and + _ 4. K. Leary, D. H. Templeton, A. Zalkin and XeOF3 Sb2Fn , LBL-I423, December 1972. N. Bartlett, The Cr/stal Structure of 2 [XeF5 ]2PdF6 -, LBL-1171, December 1972. 7. D. E. McKee, C. J. Adams, A. Zalkin, and + + Neil Bartlett, Cations KrF+, XeF3 , XeOF, , and Oxidizing Properties of KrF*\ LBL-1163, November 1972. -19- B. SOLID STATE CHEMISTRY AND PHYSICS Norman E. Phillips, Principal Investigator associated transition is cooperative. 1. HEAT CAPACITY OF SmS AT 0 AND 15 KBAR 1. K. Andres, E. Bucher, J. P. Maita, and Samuel D. Bsder and Norman E. Phillips A. S. Cooper, Phys. Rev. Letters 28, 1652 (1972J. ~ An unusual class of materials that contains rare earth ions with nonintegral valence and "soft" magnetic moments has recently been recognized.1 SmS exhibits the properties characteristic of the materials at pressures 3. THE HEAT CAPACITY OF DILUTE Cu Mn ALLOYS above 6.5 kbar, where it is metallic and the Sm ions appear to have a mixture of W. E. Fogle, Gary P. Schwartz, and 4f5 and 4f° configurations, but not at lower Norman E. Phillips pressures, where it is an insulator with the 4f* configuration. We have measured The very low Kondo temperature of dilute the heat capacity at 0 and at 15 lebar from Mn in Cu makes it possible to study the 0.3 to 20°K. Susceptibility measurements1 interactions between the Mn spins and of have shown that only the 4f" configuration the Mn spins with the conduction electrons contributes to the susceptibility, and that at temperatures at which the Kbndo spin the magnetic transition or divergence ex compensation does not occur. We have pected for the 4f5 configuration does not initiated a program of heat capacity appear. The heat capacity measurements measurements on samples with concentrations show a broad anomaly throughout the 0.3 to from 100 to 1000 at. ppm and at temperatures 20°K region that evidently extend to higher between 0.06 and 25°K. Preliminary analysis temperatures. It is clear that the entropy of the data suggests that the contribution associated with the doublet crystal field to the heat capacity of isolated impurities ground state of the 4f-> configuration is separable. disappears oveT a very broad temperature range instead of by the usual magnetic ordering transition. 4. THE LOW-TEMPERATURE HEAT CAPACITY OF 1. M. B. Maple and D. Wohlleben, Phys. Rev. NEODYMIUM Letters 27, 511 (1971). Mary M. Conway, W. E. Fogle, and Norman E. Phillips 2. HEAT CAPACITY OF PrCu2 The heat capacity of 8-cerium showed a large, weakly field-dependent, temperature- Gary P. Schwartz, M. Wun, and Norman E. proportional term that dominated the heat Phillips capacity to temperatures of the order of 0.1°K, below which it was masked by the Earlier heat capacity measurements1 on hyperfine heat capacity of impurities. To PrCu2 showed an anomaly near 6°K and the see whether similar heat capacities might high temperature side of an anomaly that be found in other rare earth metals, we peaked below 0.06°K. The 6°K anomaly was have measured the heat capacity of a interpreted as a crystal field effect that neodymium sample enriched to 99.81 in even- left the Pr3+ ions in a singlet crystal even isotopes. In zero field the heat capa field ground state, and the lower anomaly city of neodymium is similar to that of was interpreted as indicating cooperative cerium, but the field dependence is much nuclear ordering.. We have remeasured the stronger. The measurements made so far heat capacity and obtained results that are suggest tuat the linear term in the high- sufficiently different to suggest that the field limit is of the order of magnitude interpretation must be modified. The 6°K of that usually expected for the electronic anomaly is higher and sharper than a Schottky heat capacity. Tentatively, we conclude anomaly, suggesting that transition involves that the large zero-field linear terms in cooperative effects. The 0.06°K anomaly is both neodymium and cerium are probably of not as high as the earlier work showed, and magnetic origin even though the field it is therefore less obvious that the dependence is very weak in cerium. -20- 5. RESEARCH PLANS FOR CALENDAR YEAR 1973 3. T. E. Katila, N. E. Phillips, M. C. Veuro, and B. B. Triplett, Heat Capacity of Norman E. Phillips TtazCSOiiWabO Between 0.0S and 20 K , Phys. Rev. B 6, 1827 (1972) (LBL-8S8). Calorimetric investigations will be made on: (1) Dilute Magnetic Alloys. In addition to the systems currently under investigation, LBL reports Cu Mn and Au Gd, we plan to make measurements 1. M. M. Conway, The Low-Temperature Heat on (Lai. G30Al2, which is a superconductor x Capacity of Cerium Metal (Ph.D. thesis), containing localized moments, and possibly on LBL-827, May 1972. additional rare earth solutes in noble metals. There is now experimental evidence that the Kapitza resistance between 3(Ie and a metal can 2. M. P. Mathur, M. Ashkin, J. K. Hulm, be reduced by the introduction of Gd impurities C. K. Jones, M. M. Conway, N. E. Phillips, and the suitability of other rare earths might H. E. Sinun, and B. B. Triplett, Specific be predictable from heat capacity measurements. Heat of SnTe Between 0.06K and 30K Under Strong Magnetic Field, LBL-1137, August (2) Other rare earth compounds that appear 1972; To be published in the Proa, of the to show mixed configurations on rare earth 13th Intern. Conf. on Low Temp. Phys. ions, e.g., Yb Al£ Boulder, Col., August 1972. (3) Zeolites. There has been a renewed interest in the heat capacities of amorphous 3. M. M. Conway and N. E. Phillips, Low- dielectrics. Zeolites have some of the Temperature Heat Capacity of o- and B-Cerium, features of amorphous substances, voids, but LBL-1138, August 1972; To be published in the not others, disorder, and may provide some Proc. of the 13th Intem. Conf. on Low Temp. insight into the anomalous heat capacity. Phys. Boulder, Col., August 1972. (4) Liquid 4He, Rebuilding the apparatus for measurements to lower temperature will 4. R. Gardner, J. K. HoffeT, and be continued. N. E. Phillips, Thermodynamic Properties of 4He: I. The hep Phase at Low Densities, (5) Work on apparatus for temperature submitted to Phys. Rev. (LBL-1123, August scale measurements and possibly other experi 1972). ments in the mK region will be continued. 5. S. D. Bader, N. E. Phillips, and D. B. McWhan, Heat Capacity and Resistivity of Metallic SmS at High Pressure, submitted to Phys. Rev. Letters (LBL-1194, October 6. 1972 PUBLICATIONS AND REPORTS 1972). Norman E. Phillips and Associates 6. D. B. McKhan, J. P. Remeika, S. D. Bader, B. B. Triplett, and N. E. Phillips, Journals Heat Capacity of Metallic V203 at High Pressure, submitted to Phys. Rev. (LBL-1467, 1. N. E. Phillips, Low-Temperature Heat Sept. 1972). Capacity of Metals. Critical Reviews in Solid State Sciences, 467 (1971). 2. N. E. Phillips, B. B. Triplett, R. D. Clear, H. E. Simon, J. K. Hulm, C. K. Jones, and R. Mazelsky, Low-Temperature Heat Capaci ties of Superconducting Degenerate Semi conductors, Physica 55., 571 (1971) (UCRL- 19163). -21- George Jura, Principal Inveatigator was studied, and the experimental conditions changed so that the method could be completely 1. HEAT CAPACITIES AT HIGH PRESSURES generalized. In order to do this, it was necessary to remove all restrictions on the A limber of years ago, our group undertook equation used for the determination of the the development of a method for the determina heat capacity. This last enables us to deter tion of the specific heats of metals and alloys mine the heat leak constant as well as the under pressure. No calorimetric method heat capacity, and to approximate the change available at that tine was applicable for the in the ambient temperature with the passage objective, the determination of the specific of time. The results have been phenomenally heat under pressure. The method used was good. Actually, the only place where the the development of a pulse technique, but original assumptions work really satisfactorily becai*-2 of the very large heat leak, the is at one atmosphere. Indeed, it may well be work had to be done on the microsecond rather that the greatest use of the presently than the millisecond scale. Thus the entire developed technique will be its application effort has been to develop pulse calorimetry to high temperatures at one atmosphere. on this time scale. What we have found is that if the resistance Originally, we wrote the following equation of the sample is sufficiently high, then for that related the heat capacity to the quantities a period of about S00 usee the ambient can be that could be determined: considered as constant. For low-resistance material such as aluminum it is necessary to consider the change in the ambient. k(T T $"-*£-- " a>' « If the ambient is considered as constant, and it so appears for high-resistance material, where E is the voltage across the sample the integration of Eq. [1) gives when a constant current I is pulsed through the sample, RQ the resistance of the sample E-Ep = CCt-t^-tDCt-tp)2 + . . . , at the initial temperature, R' = dR/dT, k the heat leak constant, T the temperature of the sample, and T the temperature of the where E is the voltage at time t, E, the a voltage at time tR, C = AER + B, ambient. In the liirit of T=0, this reduces In to D = ATAER + B)/2. these constants, A = l2R' - k/Cp and B = kEo/Cp. In the expression EQ IS the voltage at zero time. .„ A R' If the reference voltage is taken at the center of the available interval and if the f • -rf- fo C2) time intervals are equally spaced, very little The experiiEnts then consisted of sending error is introduced in the solution for C and constant current pulses through the sample, D. In principle, k and Cp could be determined detecting the voltage drop on an oscilloscope, from their respective values when different and then determining the best value of values of the current are used in the pulse. dE/dt as t approached 0. However, this leads to an uncertainty of about 20J in the values for the two desired quantities, since two equations must be solved. There were difficulties with this technique, It is much better to approxinate EQ, which namely, the determination of 0 time. First, can be done to about 0.2S, and doing this we there is a rise time of the order of about have found that Cp and k can be determined to SO usee for the pulse, and then there is an uncertainty of about 0.3 and 1% additional ringing, leaving about the first respectively. This has been done for iron 100 usee that is not representative of the and bismuth. situation that was desired. However, encouragement was given to the In the event that the variation in the method in that when iron was investigated from ambient must be considered, we have found about 100-298°K, and from pressures ranging it satisfactory to write the original dif from 25-100 kbars, reasonable results were ferential equation of the form obtained. It was. known that the technique did not work for a metal such as aluminum. §-*-£* ^CE+0 at) , Thus, this year, the treatment of the data P P -22- where we have assumed that for the time under 30 investigation the ambient is a linear function C literature values of Fe of time, and a is the proportionality constant. p Unquestionably this is too simple, but since This work of Fe (2 mil) this correction is on the order of 10S at small tines, a large error in it has relatively small effect on the final result. This 20- treatment has been applied to aluminum, and the results are satisfactory. A word about the instrumentation. The pulse generator that we have built is satisfactory. However, the detector we are 10 using, an oscilloscope, is not the best. First, the values of the voltage obtained from reading the photograph are in error by an amount that is larger than desired. Secondly, the time required to measure the photographs to the desired accuracy is about 200 400 600 800 S hours. Time of reading, and more accurate results could be obtained with better T ft) instrumentation. For example, the use of a Fig. 1. Comparison of literature and present 12-bit transient detector would give the value values for the heat capacity of iron from room of E to 0.1$ rather than the IS we are now temperature to about 1100°K. (XBL 731-2119) getting. The appropriate instrumentation to take the data from the recorder could and would save well over 5 hours. There are other techniques besides the one mentioned that would accomplish the same purpose. The o Previous work problem is money. Any method other than that • Present work which we are using would cost a minimum of $5,000, and the most ideal about $10,000. We are going to attempt to obtain a grant for .1 1-5 this equipment. 2. HEAT CAPACITY IN IRON AND ALUMINUM 10 & 3b 4<5 5<5 & 7b 60 SlO tiff p (kbar) George Jura and Tang-hua Chen Fig. 2. The heat capacity of aluminum at Some of the data obtained by Yee in earlier 298°K as a function of pressure. work on iron were remeasured and recomputed (XBL 731-21201 using the constant ambient temperature approxi mation. The relative values of the heat capacities were the same as reported by Yee,1 tures than here reported. though the absolute values were different. It has also been possible to detect dif Iron was done at one atmosphere from 298 ferences between partially and completely to about 100°K. The results are shown in annealed iron in this kind of experiment. We Fig. 1. The solid line is from data in the believe that this is the fastest, and with literature, while the circles are the points proper instrumentation, the best available obtained in the present investigation. It is method for obtaining specific heats at high only around 1000°K that the discrepancy is as temperatures at one atmosphere. high as 2 to 3%. Undoubtedly, this discrepancy is due to the inability to obtain good values The big success of the new technique, we for dR/dT in this region. Apart from this believe, is in the determination of the heat region, the variation between the previously capacity of aluminum. Figure 2 shows the observed results is only a few tenths percent. present results at 298°K from 25-100 kbars. In the present technique the time scale was The results, within error, about 3%, indicate so short that it was not necessary to make any that the heat capacity is constant. Earlier corrections for radiation loss. There is lit work, from an approximation of the Debye tle question in our minds but that these data temperature as a function of pressure, would can be extended to very much higher tempera- predict a 1% decrease in tlie heat capacity in this range. The gTaph also shows the best Table II. results of our earlier work. p Cp/R. 1. A. F. Yee, Heat Capacity of Iron as a Function of Pressure and Temperature (Ph.D. 34.8 3.54 thesis) LBL-180, Nov. 1971. S0.1 2.90 65.4 2.72 3. HEAT CAPACITY OF BISMUTH 78.3 2.25 87.8 1.48 George Jura and Berardo Jurado One of us (BJ) has spent his time on Figure 1 is a compilation of several sets of bismuth. He has not as yet obtained dR/dT, data for the heat leak constant when AgCl is anC therefore it is essential to give only the surrounding medium. The constant rises to Cp/R1. For his studies, it is possible to a maximum around 80 kbars, and then either assume the ambient to be constant. Table I falls or levels off. There is a known phase is an example of the sort of results now being transition in AgCl at about this pressure. obtained. Hopefully, it may be possible to relate the heat conductivity of the surrounding medium Table I, Variation in the heat- capacity and to the heat leak constant. If this proves heat leak constant as a function of current at to be feasible, it will also be able to 64 kbars. determine or approximate the heat conductivity as a function of pressure, which would make some geophysicists happy. RI k/R' V C.\J i" i T- i - r 4.5 1.40 0.91 4.6 1.37 0.96 ts ) 4.7 1.37 0.98 • § • 4.8 1.31 1.10 £ 4.9 1.39 0.95 t n • 1.5 5.0 1.32 1.09 « • 5.1 1.36 1.02 S • 5.2 1.35 1.03 iC 5.3 1.33 1.07 1.0 i I i i i 20 40 60 80 100 He has also performed experiments in AgCl and a mixture of TIBr and Til mixtures as the surrounding medium. The heat conductivity of p (kbar) the thallous mixture is lower than that of the Fig. 1. The heat leak constant when silver silver chloride, and therefore should behave chloride is used as the surrounding medium, better in these experiments than the more usual as a function of pressure at 298°K. chloride. Insofar as can be told, to a (XBL 731-2121) pressure of 80 kbar, the relative values of the heat capacity are the same with either medium. 4. RESEARCH PLANS FOR CALENDAR YEAR 1973 Not all of the data have been properly treated as yet. Table II gives some of the George Jura results. There is a known phase transition between 78 and 87 kbars. This accounts for We will continue in the development of a the huge change that is listed. method for determining the specific heats -24- o£ metals and alloys under high pressures. 5. 1972 PUBLICATIONS AND REPORTS As discussed in paper 1 of this section. We hope to find the right instrumentation to George Jura and Associates record the data. We are attempting to obtain a grant for the right equipment. Journal 1. C. K. Ku, G. Jura, and M. Shen, Giiineisen Parameters of Crystalline Polyethylene, J. Appl. Phys. 43, 4348 (1972). -25- Gdbor A. Sonvrjai, Principal Investigator -i—i—i—i—i—i—i—i—i—i—r 1, LOW ENERGY ELECTRON DIFFRACTION STUDIES BY SURFACES -0.2| Benzene on PI (III) a. Low Energy Electron Diffraction and Work Function Studies of Benzene, Naphthalene, and Pyridine Adsorbed on ft till) and ft(100) -0.4 14> I0-7 torr Single-Crystal Surfaces • 1.8 x IO"T torr John L. Gland and Gabor A. Somorjai I,, • 3x 10"' torr Benzene, naphthalene, and pyridine adsorb readily on the Ft (111) and Pt(100)-(5xl) sur faces. Adsorption of these aromatic molecules I- leads to the formation of ordered structures on the Pt(lll) surface, while adsorption on the Ft (100) surface leads to the formation of poorly ordered surface layers. Benzene ad sorbed on the Pt(lll) surface at 2S°C causes ,.,, the formation of the Pt(lll)-H| ^g I -Benzene structure (Fig. 1A) with a corresponding work function change of ? -J 1 I ' i ' I i l i l l 20 40 60 80 100 120 Exposure (min) Fig. 2. Work function change as a function of exposure time for benzene on the Pt(lll) surface. The pressures should be multiplied by 5.7 to give the effective pressure at the surface. (XBL-7210-4204) approximately -1.4 V. With further benzene exposure the initial ordered structure is transformedto the PtCllll-l'sl jg I -Benzene structure (Fig. IB) with a work function change of approximately -0.7 V. The work function change as a function of exposure is shown in Fig. 2. This trans formation is the result of a reorientation of the adsorbed benzene molecules. Naphthalene forms the Pt (111)-(6x6) naphthalene structure when adsorbed on the Pt(lll) surface at 150°C, and the work function change on adsorption is Fig. 1A) A LEED pattern of the Pt(lll)-H| ^ I -Benzene structure at 29 V. B) A LEED pattern of the Benzene structure at 29 V. (XBB 7211-5893) -26- about -2.0 V. Pyridine forms two poorly c. Low Energy Electron Diffraction Studies ordered structures on the Pt(lll) surface. of Stepped Platinum Surfaces and Ordered" The large work function change on adsorption Carbon Surface Structures of pyridine (-2.5 V) indicates that the nitrogen participates in the adsorbate- Kenneth Baron, Donald W. Blakely, and substrate bond. Gabor A. Somorjai b. Low Energy Electron Diffraction and Work A series of stepped platinum crystals with Function Studies of Substituted Benzenes (111) orientation terraces and (100) steps Adsorbed on Pt(lll) and Pt(lOO) Crystal ' that differ only in the widths of the atomic Surfaces terraces have been studied. Surfaces with terraces of less than 4 atoms wide undergo John L. Gland and Gabor A. Somorjai faceting that is catalyzed by the adsorption of hydrocarbons. Stepped surfaces with Low energy electron diffraction and work wider terraces are stable and form many- function change (A*) measurements have been ordered carbonaceous surface structures upon made for a series of substituted aromatic the chemisorption of hydrocarbons. The sur compounds adsorbed on the Pt(lll) and face structure that forms depends on the type Pt(100)-(5»1) surfaces. Benzene, toluene, of hydrocarbon that is adsorbed and on the t-butylbenzene, styrene, benzonitrile, platinum surface structure. Platinum sur faces with steps of (310) orientation exhibit nitrobenzene, and aniline all adsorb on both a high density of kinks. This surface does Pt surfaces studied. On the Pt(lll) surface, not form ordered carbon structures and is not benzene forms two well-ordered structures, reactive in surface reactions. These studies are continued to explore the roie of the the Pt(lll)- "j| ^g -Benzene structure platinum surface structures in nucleating the formation of ordered carbonaceous deposits (fl*~-1.4 V) and the that appear to be important intermediates in catalytic reactions. the Pt(lll)- j| ?| -Benzene structure (A* ~ -0.7 V). d. Low Energy Electron Diffraction Studies The transition between structures is facilitated of the Clean~and Sulfur-Covered (100) and by increased exposure of the surface to ben (110) Crystal Faces of Vanadium : zene. The initial ordered structure is as sociated with a u-bonded surface species; the Clifford A. Ifegerle and Gabor A. Somorjai second with a singly dehydrogenated, o- bonded surface species. For toluene and t- Both the clean and the sulfur-covered (100) butyl benzene adsorbed on the Pt(lll)- surface, and (110) faces of vanadium were studied by the surface layer is poorly ordered, with a low energy electron diffractim (LEED) and spacing three times the underlying lattice Auger electron spectroscopy. A cleaning vectors, and is fairly exposure-independent. treatment, involving the heating of vanadium Although reorientation appears to be a major in oxygen, was devised to facilitate the pre effect for benzene on the Pt(lll) surface this paration of a clean vanadium surface. The effect is much less important for toluene and diffraction pattern obtained from the clean t-butyl benzene on the Pt(lll) surface as V(100) surface was screaked, indicative of indicated by analysis of the work function one-dimensional disorder. LEED indicates that change data. All the singly substituted both the clean aid the sulfur covered V(110) aromatic compouids studied form similar poor surfaces are ordered. The intensity of the ly ordered adsorbed layers with a characteris various dif Taction beams from the dean and tic distance three times the underlying sulfur covered V(110) surfaces were measured lattice vector on the Pt(lll) surface at 25°C. as a function of incident beam voltage. These Nitrobenzene forms the Pt(ll'l)-(4x2)-Nitro- experiments were carried out by the photo benzene structure at 100°C. graphic method. In addition, the intensity of the (00) beam from the clean V(110) surface On the Pt(100) surface, benzene forms a was measured as a function of sample tempera poorly ordered adsorbed layer with a charac ture, at several incident beam voltages. teristic distance twice the substrate unit cell size. Toluene t-butyl benzene, nitro e. Measurements of the Diffraction Beam benzene, and aniline form poorly ordered ad Intensities from the Clean Pt(100) and sorbed layers on the Pt(100) surface with a Pt(lll) Surfaces and from OrderedSurface characteristic distance three times the Structures of Uiemisoifeed Gases spacing in the underlying layer. Benzonitrile and styrene form disordered layers on the John J. Wasilezyk and Gabor A. Somorjai Pt(100) surface. A photographic technique has been developed -27- to measure the intensities of low energy elec and (111) surfaces of aluminum in order to tron diffraction beams from crystal surfaces. determine the spacing between the surface The fluorescent screen is photographed as a and bulk layers of the crystal. The Al (110) function of electron energy by a pre-sensitized surface is found to be contracted by 10 to 15% film and the uniformly developed negatives are from the bulk interlayer spacing, and the analyzed by using a micro-densitometer. The Al (111) surface is found to deviate from the integrated relative intensities of the various bulk spacing by less than 5%. This amounts diffraction beams from the clean (111) and to a determination of the surface layer posi (100) platinum surfaces and from acetylene tion to within 0.1 A, Results of calculations surface structures have been obtained this way. on all experimentally measured beams for these The intensities so obtained are compared with surfaces are compared with the experimental the calculated relative intensities obtained results for several assumed interlayer spac- by computer calculations. ings. These comparisons are made with respect to qualitative peak shapes, peak positions, f. Determination of the Surface Geometry and relative peak amplitudes of the specular for the Aluminum (1U0), (111), and (110) and all measured non-specular beams from each Surfaces by Comparison of LEEO Calculations surface. In order to achieve this agreement with Experiment it has been necessary to include the four outermost crystai layers and to describe the Mario B. Martin and Gabor A. Somorjai ion-core potential with five phase shifts in the 40-150 eV energy range. Intensity-vs.-incident electron energy (I-vs.-eV) curves are obtained for the (00)-, (11)-, and (20)-beams diffracted from the (100) 2. AUGER ELECTRON SPECROSCOFY OF surface of aluminum. The beam intensities, SURFACES calculated by using a t-matrix formalism, are sensitive to small variations in the input a. Auger Electron Spectroscopy Studies of parameters (number of phase shifts and atomic Lead-Indium Alloys' layers, order of multiple scattering, Debye temperature, incident beam angle, inner po Svent Eerglund and Gabor A. Somorjai tential models and position of the surface atoms). The use of low energy electron dif The Auger spectra of lead-indium alloys fraction for surface structure determination with 4.53, 9.08, 28.63, 44.62, and 60.49 at.% depends on the sensitivity of the I-vs.-eV lead and of pure Pb and In were monitored in curves to small variations of the surface the liquid and solid state. The relative in geometry. The reliability of this method tensities of Pb (92V) NOO Auger peak and the depends on the lack of sensitivity of these In (403V) M^ty 5N4 5 peak were used to deter calculated curves to variations of the non- mine changes in'surface composition. Deter geometrical input parameters consistent with mination of the temperature dependence of the our knowledge of the physical processes in Pb/In Auger peak intensity ratio has proved volved. We conclude that the two most sensi to be a sensitive detector of the change o£ tive input parameters are variations in the surface composition with respect to the bulk incident beam angle and small displacements of composition. The surface layers of all of the surface layer perpendicular to the sur the alloy samples were richer in lead than face. The inclusion of three layers parallel in the bulk phase. Oxygen decreased while to the surface and five phase shifts is suf the presence of carbon increased the excess ficient to obtain reliable calculated lead concentration at the surface. Both I-vs.-eV curves for the (100) face of alumi ambient gases or dissolved minute impurities num. Ihe use of a Debye temperature for the that may segregate at the surface change the surface layer which is smaller than that of two-component system to a three-component the bulk layers is shown to attenuate the system and may drastically alter the surface I-vs.-eV curves uniformly. Exclusive reliance composition. Auger spectroscopy when coupled on a peak position or on relative peak in with suitable calibration of the peak intensi tensities appears to be unjustified due to ties should be an excellent technique to unresolved uncertainties in the models for study the surface thermodynamics of multi- the ion-core potential and for the electron component systems. self-energy correction. Displacement of the surface layer parallel to the surface is shown b. AugerElectron Spectroscopy Studies of to have the greatest effect on the shape of Adsorbed Hydrocarbons on Platinum Crystal the non-(00)-beams intensity curves appearing Surfaces in the direction of the displacement. Jean Paul Deville and Gabor A. Somorjai Low energy electron diffraction (LEED) calculations have been extended to the (110) Auger spectroscopy studies of adsorbed -28- hydrocarbons on platinum surfaces have been ordered carbon-containing layer during the initiated. The purpose of these investigations reaction. is to explore the possibility of distinguish ing different types of molecules by their chemical shifts, by monitoring shifts in both +LBL-499. the platinum and carbon Auger peak positions. Preliminary studies indicate that the NNV Present address: Laboratoire de MinSralogie, platinum Auger peak at 235V shifts by 3V to Ur.iversite de Strasbourg, Strasbourg, France. 238V in the presence of crrbon deposit on the platinum surface while the KW carbon Auger b. The Hydrogenation and Oehydrogenation of peak in graphite at 270V shifts to 271.6V Cyclohexane on Platinum Crystal Surfaces for the carbon atom on the platinum surface. Donald W. Blakely, Kenneth Baron, and Gabor A. Somorjai 3. CATALYSIS BY SINGLE-CRYSTAL SURFACES The catalytic hydrogenation and dehydro- genaticra of cyclohexane has been monitored by a. Low Pressure Studies of Dehydrocyclization mass spectrometer on various stepped platinum of n-Heptane on Platinum Crystal Surfaces surfaces. The reaction path depends on the Using Mass Spectrometry, Auger Electron partial pressure of hydrogen. At low partial Spectroscopy, and Low Energy Electron pressures the reaction product is benzene Diffraction.t while at high H2 pressures cyclohexane is formed. These studies are continuing to ex Kenneth Baron, Donald W. Blakely, Bernard plore the structure sensitivity of the surface Lang,* and Gabor a. Somorjai reaction. The dehydrocyclization of n-heptane on c. Catalysis at High Pressure on Single- platinum crystal surfaces of area less than Crystal Platinum Surfaces 1 cm2 was studied in the temperature range of 100-400<>C and at pressures in the 10"* Torr Daniel R. Kahn, E. E. Petersen, and G. A. range. The toluene formation rate was moni Somorjai tored in a static system by a mass spectrom eter, the surface structure by low energy An apparatus has been constructed to study electron diffraction, and the surface composi catalysis on various faces of single-crystal tion by Auger electron spectroscopy. The CHI) platinus both in ultra-high vacuum (UHV) and face and two types of stepped platinum crystal at high pressure (1 atm). The UHV chamber faces were used. "Tie stepped surfaces exhibit consists of a 12-inch diameter stainless steel a surface structure of ordered steps of mon- vessel with a 200 liter per sec ion pump and atomic height separated by terraces with (111) a titanium sublimation pump, and is capable orientation for one face and by terraces of of reaching an ultimate pressure of (100) orientation for the other. The initial 5xl0~10 Torr. The high pressure reactor is rate of toluene formation on the stepped sur formed in the upper chamber by closing the face with (111) terraces is twice faster than knife-edge of a 3-inch i.d. stainless steel on the surface with (100) terraces, and the reactor cup onto a gold 0-ring which has been reaction can be sustained in the presence of previously seated onto the inside of the 6- hydrogen for more than an hour. A new inch reactor flange. The reactor cup is (/J * /3)-$30° surface structure which attached to the bellows drive mechanism and develops in the course of the reaction or a is capable of traversing the entiTe diameter (9 x 9) surface structure which appears upon of the reactor. The remaining five mini- heat treatment to 850°C has no detrimental flanges on the reactor flange are used to effect on the reactivity. These surface provide pressure, temperature, and concentra structures can be attributed to the presence tion measurements inside the high-pressure of ordered carbon at the platinum surface. reactor cup. A 1520 Varian Aerograph gas chromatograph equipped with dual thermal Under identical conditions, the toluene conductivity and flange ionization detectors yield of the stepped surface with (100) is used to monitor the formation of products. terraces decreases steadily as a function of the number of doses and reaches the detection Experiments to date have concentrated on limit after about sri hour. The low index (111) the feasibility of monitoring reaction products surface appears to be less reactive than the at high pressure (approx. 800 Torr) on a stepped surface with (111) terraces by at single stepped platinum crystal (Pt(s) - least an order of ma^iitude. Both of the less [7 (111) x (1111]) having a total surface reactive surfaces become covered by a dis area of 0.80 at?. Chemical reaction systems -29- under investigation are the hydrogenation of lead to the formation of additional low fre cyclopropane, the hydrogenation of propylene, quency localized surface modes that can be the dehydrocyclization of n-heptane, and the excited by the incident molecules. The ef isomerization and hydrogenolysis of neo- ficient transfer of energy via localized pentane. surface modes would increase the residence time of the incident molecules and could thus Before each catalytic run, the crystal is account for the cosine-like angular distribu 6 heated ill * 10" Torr 02 at 900-950°C for tions observed for C2H2 and CO. 120 minutes to remove carbonaceous surface residues. The crystal is then activated in a 1 atra H2 environment prior to the intro Present address: Sandia Laboratories, duction of a hydrogen/hydrocarbon reaction Livermore, California mixture. Results for iiie cyclopropane/^ system reveal that without prior H2 pre- b. Modification of Molecular Beam Apparatus treatment, there is virtually no propane product formed, whereas the reaction does Wigbert J. Siekhaus and Gabor A. Somorjai proceed at a rate very close to literature values when the crystal is pretreated in H2 The existing molecular beam apparatus was for 1 hr. modified to permit more detailed investiga tions of gas-surface reactions. The incident Work is in progress to determine the most beam path length was shortened and the pump suitable chromatographic operating conditions ing speed in the scattering chamber increased for each reaction system. Comparison of to greatly enhance the apparatus signal-to- reaction rates for each system will then be noise ratio, making possible the measurement made on various low index and stepped crystal of the scattered beam velocity. A variable- faces of Pt. frequency chopper was installed in place of the original fixed-frequency chopper, per mitting the study of surface reaction kinetics by the modulated beam technique. 4. MOLECULAR'BEAM SCATTERING BY SURFACES A. Scattering of Diatomic and Polyatomic c. H2-D2 Exchange and Helium Scattering Molecules from the (100) Crystal Face of on th"e~5teppea riatinum crystal suriace Platinum Steven L. Bemasek and Gabor A. Somorjai Lloyd A. West and Gabor A. Somorjai A mixed beam of H5 and C2 was scattered The angular distributions of thermal ener from the clean Pt-S[9(lll)x (ill)] surface at gy molecular beams of diatomic (CO, N2, O2, 900°K and HD was detected in the scattered NO. H2, D2) and polyatomic (CO2, N20, C2H2, beam. The HD was present to the extent of N02, NH,, and methylene cyclobutane) molecules 10% of the incident D2 flux. In the mixed were monitored following scattering from the beam, prior to scattering from the surface, (100) crystal face of platinum. All of the HD is present to the extent of about IS of gases studied (with the exception of ammonia) the incident Do flux, therefore the other yielded directed scattering patterns with the 9% is HD actually formed at the surface. maximum intensity peaked at or near the Previous studies of this reaction on the specular angle for scattering from a clean Pt(100) surface showed no detectable HD platinum surface or from a platinum surface formation. Work is continuing to compare covered by a layer of ordered graphite. HD foraatioi. from this stepped surface and Broad cosine-like angular distributions which the Pt(lll) surface and to elucidate the were independent of incident angle were ob reaction mechanism for the H2-D2 exchange. served, however, for the cases of C2H2 scatter ed from an adsorbed layer of C2H2 and CO The scattering distribution of a 300°K scattered from adsorbed CO. These results helium beam from the clean Pt-S(9(lll)x(lll)] seem to indicate that efficient energy trans surface was also studied. The scattering fer between the surface atoms and gas mole distribution was peaked at the specular angle cules , as indicated by angular distributions, and had intensity and angular width similar occur only in specific instances. Apparent to helium scattering distributions from the ly, the incident particles are reflected when Pt(100) crystal surface. The scattering ever the non-localized vibrational energy intensity is much lower and the distribution modes of surface atoms and the translational broader, for the stewed surface of this energy of the incident molecules are of simi study, than that of the helium scattering lar magnitude. The adsorption of gases may distribution from the Pt(lll) crystal surface. -30- 5. RESEARCH PLANS FOR CALENDAR YEAR 1973 Using Mass Spectrometry, Auger Electron Spectroscopy and Low Energy Electron Diffrac Gabor A. Somorjai tion, J. of Catalysis 27, 40S (1972) (LBL-499). Low energy electron diffraction studies of 7. G. A. Somorjai, Principles of Surface iron and molecular crystal surfaces (ice, Chemistry (Prentice-Hall, 1972). napthalene, etc.) will be initiated. Surface structure analysis and computer calculations B. G. A. Somorjai, The Surface Structure of will be carried tut using vanadium and platinum and Catalysis by Platinum Single Crystal single-crystal surfaces and surface structures Surfaces, Catalysis Rev. 7_, 87 (1972) (LBL-455). of adsorbed gases. 9. F. J. Szalkowski and G. A. Somorjai, Auger Catalytic reaction studies on single- Electron Spectroscopy Investigations of the crystal surfaces at low pressures and at high Surface Chemical Composition of Vanadium, the pressures will be extended to include the Vanadium Oxides, and Oxidized Vanadium: Chemi cyclization of neopentane, hydrogenation of cal Shift and Peak Intensity Analysis, J. Chem. benzene, and dehydrogenation of cyclohexane. Phys. 56, 6067 (1972) (LBL-444). Auger electron spectroscopy studies of hydro 10. L. A. West and G. A. Somorjai, Scattering carbons adsorbed on platinum surfaces will be of Diatomic and Polyatomic Molecules from the carried out. (100) Crystal Face of Platinum, J. Chem. Phys. 57., 5143 (1972) (LBL-895). Molecular beam scattering studies of sur w 11 De car face reaction of H?, L>2 > °z ^ ried 11. G. A. Somorjai, On the Mechanism of Sulphur out by measuring the velocity of the scattered Poisoning of Platinum Catalysts, J. of beam. Catalysis 27, 453 (1972) (LBL-8S9). 12. G. A. Somorjai, R. W. Joyner and B. Lang, The Reactivity of Low Index [(111) and (100)] 6. 1972 PUBLICATIONS AND REPORTS and Stepped Platinum Single Crystal Surfaces, Proc. R. Soc. Lend. A. 331, 335-346 (1972). Gabor A. Somorjai and Associates (LBL-863). Journals and books LBL reports 1. L, A. West and G. A. Somorjai, Reactive 1. M. R. Martin and G. A. Somorjai, A Systema Scattering of N2O Molecular Beams from Clean tic Evaluation of the Sensitivity of Leed and Carbon Covered (100) Platinum Single Calculations to Small Variations of the Input Crystal Surfaces. Proc. Conf. on Solid Surf., Parameters for the Aluminum (100) Surface, J. Vac. Sci. 5 Tech. 9_, 668 (1972) (LBL-133). LBL-n.,4, August 1972. 2. M. Fink, M. L. Martin, and G. A. Somorjai, 2. Mario Martin, A Computer Program to Calcu Comparison of Backscattering Intensities for late Low Energy Electron Diffraction Intensi Low Energy Electrons from Various Surface ties, LBL-1418, December 1972. Atoms, Surface Sci. 29, 303 (1972) (LBL-144). 3. Steven L. Bernasek and G. A. Soirorjai, 3. F. R. McFeely and G. A. Somorjai, Studies Nozzle Beams, A Source of Molecules with of the Vaporization Kinetics of Hydrogen Bond Intriguing Properties, LBL-1140, Novenfeer 1972; ed Liquids, J. Phys. Chem. 76, 915 (1972) submitted to J. Chem. Education. (LBL-127). 4. J. L. Gland and G. A. Somrjai, Low Energy 4. B. Lang, R. W. Joyner, and G. A. Sonorjai, Electron Diffraction and Work Function Studies Low Energy Electron Diffraction Studies of of Benzene, Naphthalene and Pyridine Adsorbed High Index Crystal Surfaces of Platinum, on Pt(lll) and Pt(100) Single Crystal Surfaces, Surface Sci. 30, 440 (1972) (LBL-121). LBL-1422, December 1972; submitted to Surface Science. 5. B. Lang, R. W. Joyner, and G. A. Somorjai, Low Energy Electron Diffraction Studies of 5. G. A. Somorjai, Low Energy Electron Dif Chemisorbed Gases on Stepped Surfaces of fraction and Auger Electron Spectroscopy Platinum, Surface Sci. 30_", 454 (19721 (LBL-175). Studies of the Structure of Adsorbed Gases on Solid Surfaces, Conference on Surfaces, 6. R. W. Joyner, B. Lang, and G. A. Somorjai, University of Missouri, Rolla, June 19-23, Low Pressure Studies of the Dehydrocyclization 1972; to be published in Surface Science, of n-heptane on Platinum Crystal Surfaces LBL-878, June 1972. -31- Lee F. Donaghey, Principal Investigator concentration for a given thermal gradient, because of the variation of equilibrium 1. GAS-SOLID PROCESSES constants with temperature, as shorn in Fig. 1. a. Chemical Transport Reaction Analysis for Crystal Growth of Binary and Ternary The chemical transport of ternary chalco Chalcogenides* genides with halogens as transport agents shows similar rate dependence on thermal Rajani B. Agnihotri and Lee F. Donaghey gradient and transport agent concentration, but the increased number of gas phase species The chemical transport reaction method raises the complexity of the simulation of crystal growth in closed systems w<<5 problem. analyzed for binary and ternary (AB2X4) chalcogenides (X = S, Se, Te). The Stefan- Maxwell equations for multicomponent diffu Abstracted from LBL-1404 (M.S. thesis). sion were integrated to obtain concentration profiles of gaseous species along the b. Chemical Vapor Deposition of GaAsi- Px , transport path. Gas phase diffusion con x trolled kinetics were assumed to apply in Reactor Design and Growth Kinetics* the range of pressure between 10-3 and 3 atmospheres and fur temperatures between Saleem A. Shaikh and Lee F. Donaghey 800 and 1500°K. A general solution for a The chemical vapor deposition of GaAsj. P set of simultaneous nonlinear equations was x x obtained by unconstrained minimization from GaU), PH3, ASH3, HC1 and H2 source techniques. With chemical equilibrium chemicals in a flow-through reactor was assumed at solid-gas interfaces the simula tion allows the prediction of the transport do ZOHE REACTION ZONE PEPOSinON ZONE rate, the concentration profiles of gaseous species along the transport path, and the total pressure. Studies of the chemical transport of CdS, ZnS, and MnS with iodine show that the transport flux typically increases with the thermal gradient and concentration of the transport agent. For transport of MnS with iodine, however, the product flux reaches a maximum with respect to the iodine feed T 1125*—IJOO*K OOU-i 1 L 1 0091 2 3 10 20 lodina Fata ImoKi/cm'luO* Distonc« in rhe Reactor, in. Fig. 1. MnS product flux chemically trans Fig. 2. Schematic dependence of gas phase ported fcr fixed temperature gradients as a partial pressures in lihe gallium saturation, function of iodine feed concentration. reaction, and deposition zones of the (XBL 7212-7344) GaAsi-xPx reactor. fXBL 729-7010) -32- studied both theoretically and experimentally. of the group V elements whenever the group Reactor designs and reactions were reviewed, III element is in excess of the stoichiometric phase equilibria were calculated from known amount required. reaction equilibria, and diffusion limited deposition kinetics were formulated. d. Reactive Sputtering Thermodynamics of phase equilibria in the Ga-As-P-Cl-H system were studied for each Conrad Wright and Lee F. Donaghey of three zones of an experimental reactor, and deposition kinetics calculated based A sputtering system is currently being on a boundary layer model. constructed with the capabilities of dc and rf sputtering. The sputtering system will A schematic of the gas phase species con be used to explore reactive sputter deposition centration in the gallium saturation, of magnetic oxide compounds, electrically reaction, and deposition zones of the reactor conducting electrode materials for applica is shown in Fig. 2. The mole fraction x of tion in liquid crystal display devices, and GaP in the deposited solid solution and the compound catalysts. Sputtering rates will deposition rate were computed as a function be studied, in addition to the effects of of input concentrations and temperature changes in film stoichiometry with reactive profile, and studied experimentally. The gas pressure and target composition. mole fraction of GaP in the solid solution was found to increase with substrate tempera ture with a rate dependent on several 2. CRYSTALLIZATION PROCESSES factors. The deposition rate, however, decreases with increasing substrate tempera a. Forced Convection During Czochralski ture and is relatively independent of input Crystal Growth molar ratios. A critical HQ input was found for maximum utilization of the group V Pei Shiun Chen and Lee F. Donaghey elements in the reaction in each of the dif fusion controlled and surface reaction con Forced convection under conditions of trolled regimes. The experimentally obtained Czochralski crystal growth was studied in the isothermal limit. In this method of alloy compositions of GaAsi_xPx deposited on [100] Ge and [100] Si substrates agree crystal growth the liquid phase hydrodynamics well with predicted values, but the experi significantly affect inhomogeneities in dis mental gTowth rates were far below the location densities and dopant concentration expected rates. in the growing crystal. The forced convec tion process was simulated experimentally at low temperature by using a transparent crucible and simulation crystal with ethylene Abstracted from LBL-1149 (M.S. thesis). glycol solutions as the working fluid. The simulation revealed several vortices whose c. Chemical Vapor Deposition Rate Analysis surfaces of separation could be displaced of Group 1II-V Compounds by varying the rotation speeds and direction of the simulation crystal and crucible. When Ronald H. Bissinger and Lee F. Donaghey both rotated in the same direction [co-rotation) one stagnation surface separates Factors affecting the deposition rate of two non-intermixing regions preventing com chemical vapor deposition reactions involving plete bulk liquid homogeneity. During group III-V compounds are under study for counter rotation two stagnation surfaces gas phase diffusion and mixed diffusion- separate three non-intermixing regions. surface reaction kinetic control. The Crystal rotation controls the convection primary rate process neglected in previous above the new stagnation surface and the studies has been the effect of gas phase rate of mass transfer across stagnation reaction within the thermal boundary layer surfaces. The study is currently being adjacent to the substrate. Under high thermal extendeu to the non-isothermal case and to gradients the shift in reaction equilibria experimental studies of Czochralski growth with temperature decreases the deposition of intermetallic and ionic compounds. rate. The present study improves on previous analyses by including the mixed polymers of b. Segregation of M1CI2 in NaCl During the group V elements', AsP, AS2P2, AS3P, and Czochralski Crystal Growth ASP3. The equilibrium constants for reac tions in which these species participate, Nguyen Trien and Lee F. Donaghey estimated from statistical thermodynamics, should be applicable to gas phase equilibria Nonhomogeneous segregation of MnCl2 in for any III-V compound alloy. The deposition NaCl during Czochralski crystal growth were rate analysis must include the mixed polymers studied by electron diffraction, Raman -33- spectroscopy and internal oxidation. Divalent by Raman scattering since the impurity removes cation solubility is limited in alkali inversion symmetry in the NaCl lattice, halides but dependent on growth rate making first-order Raman scattering possible. transients. Thus, variations in crystal Several scattering frequencies were recorded growth rate will produce double layers con in the range from 115 to 265 cm"!. sisting of enhanced incorporation of + (Mn2 , VNa+) and reduced incorporation, c. Electrodeposition of Sodium Tungsten relative to the average solid concentration Bronzes of Mn2+. The enhancement bands can be observed by precipitating MnClj at tempera- Calvin T. Chew and Lee F. Donaghey tuxesbelow the growth temperature, and by internal oxidation. The internally oxidized The single-crystal growth of sodium tung precipitate layers were studied in sten bronzes with the general formula Na^KC^, Czochralski-grown crystal sectioned along 0.2 < x < 0.9, has been studied by electro- the axis of growth. deposition at varying rates from molten mix tures of sodium tungstate and tungsten triox- Dark field electron microscopy revealed ide. These electrically conducting compounds precipitates in aged specimens of l-10u size have been recently suggested* uz electrodes on numerous crystallographic planes. The in alkali fuel cells where high electrical MnCl2 substitutional impurity was detected conductivity and resistance to corrosion and dissolution are required. In this study a single cell with a spherical cathode electrode was employed in crystal growth at constant current or at constant overpotential. Surface morphologies of crystal growth source areas were examined by scanning elec tron microsccpy. A multiple screw dislocation growth source is shown in Fig. 1, for the com pound Nan 8WO4. In general the density of lay er edges increased with electrolysis current. Dendritic breakdown of layer edges occurred at high current, accompanying an incorporation of secondary phases in the growing crystal. 1. D. B. Sepa, A. Damjanovic, J. O'M. Bockris, Electrochim. Acta 12_, 746 (1967). d. Orthoferrite and Garnet Synthesis and Crystal Growth Conrad Wright and Lee F. Donaghey Rare earth orthoferrites of SmFe03 and TbFe03 were produced by hot pressing the oxide powders at temperatures up to 1400°C and 7000 psi for 104 seconds. Examination of the specimens by electron microprobe analysis re vealed nearly complete reaction of the primary oxides. In many samples, however, traces of unreacted R2O3 and Fe203 in the hot pressed materials were observed, indicating that higher temperatures and pressures are necessary for complete reaction. Similar results were obtained on hot-pressing the rare earth garnets SnrjFesO^ and Th3Fe50i2. Successful synthesis of SmFe03 was real ized by a zone refining technique following partial reduction of the compound. S\nce the Fig. 1. A multiple screw dislocation growth electrical conductivity of the stoichiometric source in Nao gWO^ produced by fused melt constituent phases was not high enough to electrolysis.' (XBB 731-86) permit electron beam zone refining, an inter- -34- mediate arc melting process was introduced initial liquid and substrate temperatures and found essential for tl.is synthesis. differ, and for the temperature-dependent The orthoferrite oxides were first arc-melted case where the substrate-liquid system bound into a polycrystalline rod at low oxygen ary temperature is programed. Conditions partial pressures which then, because of the for which heat transport can be included in increased electrical conductivity, could be mass transfer controlled solidification from zone refined with electron beam heating. solution are currently under study.for several nrocesses. e. Liquid Phase Epitaxy of Magnetic Garnet Films Conrad Wright and Lee F. Donaghey g. Ternary Laminar Eutectic Morphology Isothermal liquid phase epitaxy has been Lee F. Donaghey employed for the purpose of producing uniform thickness low-defect-density magnetic garnet The Cd-Sn-Pb ternary eutectic reaction is films on non-magnetic substrates. This method unique in exhibiting a cooperative ternary is based on the observation that melts of laminar structure in which the Cd- and Pb-rich appropriate flux to garnet conpositions can phase lamellar are separated by the Sn phase. be supercooled over a temperature range, thus The solidification-dependent morphological permitting epitaxial growth of the garnet properties of ternary laminar eutectics in films at constant temperatures. Melt com general and for the Cd-Sn-Pb eutectic in positions were composed of stoichiometric particular have been studied toward pre quantities of the rare earth oxides and an dicting microstructural properties from excess of Fe203 dissolved in a 50:1 weight liquid phase diffusivities and interfacial ratio of PbO:B203 flux. Melts have been free energies. The Cd-Sn-Pb eutectic studied with oxide to flux ratios ranging morphology predicted from minimizing the from 4 to 14 mol %. Figure 2 shows a dynamic system free energy is consistent simplified garnet ternary system showing with that observed experimentally. Measure the various oxide and flux ratios used in ment of interlamellar spacings, X, in the liquid phase epitaxial studies. The eutectics solidified at varying velocities, liquidus temperature was determined by v, give the relation, gadolinium-gallium gamet (GGG) substrate immersion probes at various temperatures for each melt composition studied. X2v =21.2 um3/sec , Fe,0 3 a fundamental relation or this microstruc- Gornet oxide ratioJY\ ture. in melt ~^—y \ \ Mol % Garnet^L Orthoferrite^oN. V1. Approximate region S. \v °* epitaxial growth of ^^ 3. SOLID ELECTROLYTE STUDIES from earlier studies, particularly those using "open cell" techniques, on the above a. Design Criteria for Solid Electrolyte systems is higher than that measured in this Electrochemical Cells" study. Incorrect temperature measurement due to thermal gradients around the electrolyte Richard W. Headrick and Lee F. Donaghey was also shown to be a source of error. Factors affecting the accuracy of galvanic cells involving solid electrolytes in thermo- Abstracted from LBL-839 (M.S. thesis), chemical measurements were studied with July 1972. regard to cell design and operation. Because coexistence electrodes are particularly well b. Activity Measurements in Liquid Metal suited for determining free energies of Alloys Using Solid Oxide Electrolyte" formation of metal oxide compounds, the cells Techniques Ray Pong and Lee F. Donaghey Cu, Cu2O|Zr085 Ca015 O^lM, NiO The thermodynamic properties of liquid metal alloys are under study by solid oxide and electrolyte galvanic cell techniques. Activities of group III elements in liquid Fe203, Fe3O4|Zr085 Ca0a5 O^lNi, NiO alloys have been studied with both ^0.85^.1501.85 CCSZ) and Tho.g5Y.15Ox.925 (YDT) solid electrolytes in the low oxidation exhibiting low and high entropies of reaction, potential region where non-equilibrium respectively, were examined for critical processes occur. Instrumentation for the parameters. continuous monitoring of cell potential was constructed and its limitation explored. The galvanic cells studied were based on Component activity measurements in the In-Ga the closed cell concept involving sealed system utilizing the cell Pt|W| In:Ga = electrode compartments. Design improvements Ga203 ||{YDT, or CSZ}|| Ga:Ga20j|W|Pt show include (1) interior thermocouples for a time-temperature dependent decline in cell accurate measurement of the electrolyte emf consistent with the finite electronic interface temperature, (2) individual gas conductivity of the electrolytes. Properties and vacuum manifolds on the isolated half- of electrode reactions and of electrolytes cell compartments, (3) elimination of stray are being studied to correct for non- emf in potential leads, and (4) tapered equilibrium processes affecting activity alumina to glass wax seals to facilitate measurements in liquid metals at low oxygen rapid electrode replacement and resealing. partial pressures. The improvements introduced in galvanic The conditions under which solid oxide cell structure led to standard free energy electrolyte techniques can be applied to change determinations for the following thermodynamic property measurements of III-V reactions • compound alloys have been defined for each of several solid electrolytes. The range of allowable temperature and oxidation potential Ni(s) + Cu2°(s)"Ni0(s) + 2Cu(s) > was determined for CSZ and YDT solid electro lytes by utilizing the most stable oxide and AG" = - 15858 + 3.1140T ± 8 cal element as the reference electrode. The high vapor pressures of the group V elements Ni ,s, + 3Fe203 *• 2Fe304 + NiO imposes further restrictions requiring high pressure cells over some ranges of measure AG0 = 3641 - 13.963T ± 29 cal . ments. c. Defect Structure of Gadolinium-Gallium The most significant error in cell voltage Garnet (emf j measurements was found to arise from secondary electrode reactions with water Conrad Wright and Lee F. Donaghey vapur. Proper outgassing of the half-cells was determined to be e:ctremely important, as The non-stoichiometric defect structure even a slight amount of retained moisture of gadolinium-gallium garnet (GGG) is being caused higher cell emfs. Secondary electrode determined by electrical conductivity methods reactions with retained water vapor are proposed as a mechanism whereby emf data over a temperature range of 800-1500°C and from 1 to 10"" atm of oxygen. The electrical -36- conductivity is measured by a standard dc preparation and processing of solid state four-point probe technique using Pt electrodes. materials and the impact of processing on The specimen is enclosed in an air-tight physical properties. Kinetics are emphasized alumina cell with provisions for passing in crystallization processes and in deposition gas through the cell. from the gas phase. Epitaxial thin films of compound semiconductor alloys and magnetic A gas-mixing manifold was designed to oxide compounds will be studied through produce the desired oxygen partial pressures phase equilibria, defect structure, and in the cell by appropriate dynamical mixing processing effects. of CO and CO2 gases. The manifold is capable of mixing the gases in a 1:100 to 100:1 Gas-phase etching kinetics of GaAs-GaP ratio by the proper choice of orifices. alloys will be studied. Group V polymer Figure 1 shows the conductivity of GGG over formation will be examined and related to a temperature range of 1000-1500°C for the chemical vapor deposition process. X-ray CO:C02 ratios of 47.4 and 0.8S. topography will be employed to study the dislocation structure in graded-composition The conductivity method is currently being epitaxial layers. extended to magnetic thin films epitaxially deposited on GGG to determine defect structure Magnetic oxide compounds are under study of these films. by several techniques. An rf sputtering module is nearly completed, and will be used to reactively sputter orthoferrite and garnet compounds. Alternative solvents to PbO will be examined for preparation of garnet com 1 atm 0, pounds by liquid phase epitaxy, and the deposition rate studied for simultaneous C0/C02 - 0.85 CO/CO- - 47.4 . heat and mass, transport. Magnetic properties and defect structure of garnet films appli cable in magnetic bubble domain devices will be studied. Phase equilibria in ternary liquid alloys of indium and gallium will be studied by solid electrolyte electrochemical cell techniques. Activity measurements are lacking for many group III-Va-Vb and group IIIa-IIIb-V compounds, yet can be measured directly by solid electrolyte galvanic cells. Dopant segregation during Czochralski crystal growth will be studied in CaF2. Metal-intermstallic eutectics will be grown by this method to explore thermal instabilities, thermal profiles and the effect of forced convection. Electrodeposition studies of alkali tungsten bronzes will be directed toward measuring potential distributions and toward 1/TX104 ("K ) deducing the melt complexes taking part in Fig. 1. Electrical conductivity of crystal growth. gadolinium-gallium garnet as a function of temperature and oxygen partial pressure. Phase equilibrium measurements of nematic (XBL 731-5645) liquid crystals containing pleochroic dyes by differential scanning calorimetry and optical scattering will continue toward 4. RESEARCH PLANS FOR CALENDAR YEAR 1973 determining temperature ranges over which electronic color switching and dynamic Lee F. Donaghey scattering can be obtained. Application of these effects to optical display devices This program will continue to concern will continue to be explored. physiochemical methods involved in the -37- 5. 1972 PUBLICATIONS AND REPORTS S. L. F. Donaghey, Applications of Gas-Solid Reactions to Materials Coating, Chemical Lee F. Donaghey and Associates Engineering Colloquium, University of California, Berkeley, Dec. 6, 1972. Journals LBL reports 1. S. Bjorklund, L. F. Donaghey and M. Hillert, The Effects of Alloying Elements 1. R. W. Headrick, Design Criteria for Solid on tiie Rate of Ostwald Ripening of Comentite Electrolyte Electrochemical Cells (M.S. in Steel, Acta Met. 20, 867-874 (1972).* thesis), LBL-839, July 1972. Papers presented 2. S. A. Shaikh, Chemical Vapor Deposition of GaAsi-xPx, Reactor Design and Growth 1. I. F. Donaghey, Selected Topics in the Kinetics (M.S. thesis), LBL-1149, September Thermodynamics and Kinetics of Alloys with 1972. Reference to Steel, OAS Lectureship, Institute of Physics, National University of 3. R. A. Agnihotri, Chemical Transport Mexico, Mexico City, June 12-30, 1972. Reaction Analysis for Crystal Growth of Binary and Ternary Chalcogenides, (M.S. 2. L. F. Donaghey, Direct Observation of thesis) LBL-1404, December 1972. Phase Transformations with High Voltage Electron Microscopy, ibid., June 20, 1972. 4. P. S. Chen and L. F. Donaghey, An Experi mental Study of Forced Convection during 3. L. F. Donaghey, Applications of Solid Czochralski Crystal Growth LBL-1432, Electrolyte Galvanic Cells in Inorganic December 1972. Materials Research, ibid., June 27, 1972. 5. R. W. ifeadrick and L. F. Donaghey, Design 4. L. F. Donaghey, Materials Research at Criteria for Solid Electrolyte Electro Berkeley, Institute for Advanced Studies, chemical Cells, LBL-839 Rev., Decenber 1972. Polytechnical Institute, Mexico City, June 30, 1972. Supported by the Swedish Agency for Technical Development. -38- EUECTR0CHEM1STBY Charles W. Tobiae> Principal Irweatigatov Accuracy in this region was obtained by using narrow cathode segments which enabled measure 1. FUNDAMENTAL STUDIES OF IONIC MASS TRANSPORT ments of local mass transfer rates. Figure 1 shows the ratio of Nu/Nuw as a function of a. Distribution of Ionic Mass Transport Rates distance along the plate expressed in equiva in a Flow Channel' lent hydraulic diameters. As expected, the mass transfer entrance region narrows down Uziel Landau and Charles IV. Tobias with increasing flow rates; here the local Nusselt number is proportional to the inverse Studies of the local maximum mass transpo-.-t cube root of the distance from the leading rates along planar electrodes in channel flow were extended to the turbulent regime. Local transport rates were measured by using the limiting current technique for copper deposi 1. Uziel Landau and C. W. Tobias, Flow Circuit tion from acidified copper sulfate solutions for Ionic Mass Transfer Studies at High Flow on a segmented copper electrode embedded in Rates, IMRD Annual Report for 1970, UCRL- the wall of a flow channel. The pilot plant 20S00, April 1971, p. 39. scale flow circuit and its instrumentation were reported earlier.* * b. Mechanism of Electropolishing of Copper Pressure-distribution measurements taken in the flow channel in the turbulent flow Kaoru Kbjima and Charles W. Tobias regime prove that a fully developed hydro- dynamic turbulent flow profile has been Diffusion kinetics of the active dissolution established. Well-defined limiting current of copper in phosphoric acid was studied under plateaus were obtained up to Reynolds number both galvanostatic and potentiostatic condi of 100,000 (corresponds to linear velocity tions. In agreement with Elmore, and Hickling of 300 cm/sec). The power dependence of et al., the first voltage rise in Nusselt number on the Reynolds and Schmidt galvanostatic experiments, as well as the numbers obtained experimentally is in agree onset of current decrease in potentiostatic ment with existing correlations for mass experiments, is interpreted to be caused by transport in turbulent flow. exceeding the solubility limit of copper phosphate at the anode surface. The potential or current value respectively can be predicted Special attention was paid to the entrance by evaluating the rate of transport of copper region where mass transfer rates are higher. phosphate away from the surface. Re-examination of published experimental data, as well as numerical evaluation of a newly proposed theoretical model, indicates that in electropolishing the concentration of phosphoric acid at the anode surface is far from zero: pH *> 2. Correction of literature values of anodic 1 \ Re =80,00 0 overpotential for ohmic potential drop between anode and reference electrode and for con Re • 50,000 centration over-iotential reveals that in the « \ active dissolution range the behavior of charge transfer overpotential can be well Re « 30,000 represented by the Tafel relation; the cal Vf \ culated value of a is similar to that reported S. Re » 5000 a for dissolution of copper in acid sulfate media. ^ i i i i i The characteristic current peak separating 4 6 the active dissolution region from the current plateau, where polishing occurs, is shown to x/D h result from increasing coverage of the anode Fig. 1. Experimental mass transfer rates surface by cuprous oxide with increasing in the entrance region. (XBL 731-2144) applied potential. -39- Thermodynamic and transport calculations lead to a coherent picture of the stability of cuprous and cupric oxides in the brightening-polishing range. A reinterpreta- tion of impedance data obtained by Qhashi et al. on oxide-covered copper anodes indicates that the range of thickness of the invisible layer responsible for the polishing action is 13 - 124 A. Optical observations, thermodynam ic calculations, and impedance data indicate that in the polishing region the oxide covering the surface is multilayered, prob-^v composed of layers of CU2O, CuO, and a higher oxide of copper. A method is presented for the prediction of the current density range suitable for electropolishing under various hydrodynamic conditions. * Abstract of LBL-140, Feh nary 1972. 2. ELECTROLYSIS AT HIGH CURRENT DENSITIES * I i i I I . I I ' ' a. Mass Transfer at Closely Spaced Planar 2 5 2 5 2 5 Electrodes lO-2 10-' 1.0 10 Interelectrode gap (cm) Raul E. Acosta, Rolf H. Muller, and Fig. 1. Power required for pumping electro Charles W. Tobias lyte between parallel olanar electrodes. J _z 2 _1 The program to establish working correla Based on: p = 1 gcnf , v = 1 Kl0 cm sec , tions for mass transfer processes under the D = 4*10-6cm2sec-l. (XBL 728-3731) conditions found in high current density electrolysis has been continued. These conditions, involving the use of a very small for mass transfer data that appear in the interelectrode gap (of the order of a few literature: the data are often reported in hundredths of a centimeter) and high flow terms of a fraction factor for "smooth" elec rates, give rise to prohibitively large trodes , even though this friction factor was pumping requirements. However, for the same net determined for the particular equipment mass transfer rate fsame mass transfer boundary where the mass transfer data were obtained. layer thickness 6j^) the power requirement decreases with decreasing interelectrode gap (Fig. 1), making the use of small gaps less Limiting current measurements have also unattractive for large scale processes. been performed in the range 1,000 < Re < 20,000 using three different channel widths (1, 1.5, By using integral pressure transducers, the and 2 cm) and an interelectrode gap of 0.05 cm pressure drop along the channel cell described (Fig. 2). The results confirm that for an earlier* was determined in the range aspect ratio (width/gap) greater than 20 the 1,000 < Re < 20,000, to obtain the friction system may be considered to be formed by two factor. These pressure measurements confirmed parallel plates of infinite width. that the entrance length allowed (SO DhO was sufficient to insure a developed velocity A new cell and pneumatic pump have been profile at the electrodes. Even though the installed and tested. This new system will cell walls had been polished to optical allow the study of mass transfer rates using smoothness (the roughness measured with a an interelectrode gap of 0.02 cm, in the same profilometer was of the order of 0.5 y), Reynolds number range. Roughness effects will the roughness factor for this cell is high be studied for both 0.05 and 0.02 inter (0.0005), since the interelectrode spacing is electrode spacings, using a random sandblast only 0.05 cm. This points out one of the finish and a geometric V groove pattern uncertainties involved in using corT3lations perpendicular to the flow direction. -40- 1000 7 " i ~ 1 | 1 1 11 f —1 responsible for the favorable properties . displayed by the sodium chlorate electrolyte - 0_ R. = ,A*0 ' in electrochemical machining applications.5 - - °-———^_Sj600 Galvanostatic dissolution of copper in - - sodium chlorate electrolytes occurs by ionic conduction of cuprous ions through a thin - - uniform oxidic film which forms on the surface of the metal. This film gives rise to a """-^ 4,000 uniform surface impedance which results in a generally polished surface texture. " ^ ^ ^^2,000 The upward sweep in potential is believed "*•"«* to be caused by the restriction of current 100 on the anode film by the precipitation of a - *-^ ~ secondary layer of cuprous oxide and basic ^-» copper salts. At a given current density, the - " rate of accumulation of the secondary layer is controlled by the local concentration of hydroxyl and chloride ions and by the rate of Moss Transport to a Planar Electrode convection of cuprous ion away from the - - surface. Intcrelectrode 6ops0.05cm The periodic rapid decreases in anode potential may be the result of a dielectric breakdown of the uniform oxidic film followed by a short period of active dissolution of 10 .1 1 1 1 1 1 1 1 1 1 10 the substrate. The abrupt fall in potential is accompanied by a spark discharge at the x/Dh anode surface, followed by a rapid etching of Fig. 2. Mass transport to a planar electrode. the metal surface. Interelectrode gap = 0.05 cm. Electrolyte properties same as used in Fig. 1. It has been found that similar oscillations (XBL 728-3761) may be produced in a variety of electrolytes (e.g., molar sodium nitrate, -sulfate, and -hydroxide) by the addition of small amounts of halide ions. Experiments an under way to Surfanalyzer 150 System, Clevite Corporation, relate the oscillations to the widely reported El Monte, Calif. phenomenon" of halide ion attack on anodic oxide films. 1. Raul E. Acosta, Rolf H. Muller, and Charles W. Tobias, Electrolysis at High Current Density: Mass Transfer Considerations, IMRD Annual Report 1971, LBL-425, April 1972, p. 35. 1. K. J. Vetter, Electrochemical Kinetics b. Periodic Phenomena in the Anodic (Academic Press, New York, lHbVJ, p. /Bo. Dissolution of Copper 2. R. S. Cooper and J. H. Bartlett, J. Electrochem. Soc. 105, 109 (1958). John F. Cooper, Rolf H. Muller, and 3. J. Cooper, R. FTTMuller, and C. W. Tobias, Charles W. Tobias "Periodic Phenomena in Copper Dissolution at Constant Current," paper presented at 138th Oscillations in anode potential have been meeting of the Electrochemical Society, observed during galvanostatic dissolution with, October 4-9, 1970. a variety of metal/electrolyte combinations.1'2 4. D. Jacquot, ECM Studies of Engineering Of particular interest aro the regular saw Materials in a Flow Channel Cell, to be sub tooth potential oscillations which occur mitted for publication. during the high current density dissolution 5. M. A. LaRoda and M. C. McMillan, Electro of copper^ and low carbon steel* in sodium chem. Tech. S, 340 (1967). chlorate solutions. Anode potential oscilla 6. B. N. KaEanov, Electrochemistry of Metals tions have been related to the periodic growth and Adsorption (Freuna Publishing House, and rupture of anodic oxide films3 and may be Kolon, Israel, 1969) p. 155. -41- 3. ELECTROCHEMISTRY IN NONAQUEOUS SOLVENTS 4. RESEARCH PLANS FOR CALENDAR YEAR 1973 a. Electrochemical Behavior of the Alkali Charles W. Tobias Metals in Propylene Carbonate* a. Fundamental Studies of Ionic Mass Transport Jacob Jorne" and Charles W. Tobias This program is designed to develop under The electroreduction of active metals from standing of the effects of hydroctynamic flow aqueous solution is restricted primarily by on electrode processes in models, which have thermodynamic limitations, because of prefer clear-cut significance in applied electro- ential decomposition of the solvent. The most synthesis, batteries, and fuel cells. Emphasis reactive of these, the alkali metals, cannot will continue to be on flow between parallel be deposited at all from aqueous solutions. planar electrodes. Integral limiting currents Propylene carbonate (PC), a highly polar and distribution of limiting rates are aprotic compound, was used as solvent medium measured in a systematic manner so as to for the deposition of the alkali metals at enable us to propose correlations of general ambient temperature. Li, Na, K, Rb, and Cs validity over broad ranges of Reynolds and were deposited successfully from their Schmidt numbers. In addition, investigations chlorides in AICI3 solution in PC. AICI3 will be extended to observe effects of flow on served as a complexing agent. The experimental surface morphology during the electrodeposition program included: of metals. A parallel effort will be conducted 1. Equilibrium potential measurements pertaining to the direct measurement of con which yielded activity coefficient data, and centration gradients in the solution boui.Jary served to establish a relative standard layer by Mach-Zehnder interferometry (see potential scale in AICI3-PC. project under R. F. Muller). Effective ways 2. Conductance and density measurements, for the reduction of the boundary layer 3. Micropolarization measurements, for thickness by means other than bulk flow will obtaining kinetic information (exchange current be sought. These may include slurry flow and densities, transfer coefficients) on the alkali triggering entrance effects by periodic local metals and their amalgams in AICI3-PC. interference with the boundary layer. 4. Electrodeposition studies in the Tafel region. b. Electrolysis at High Current Densities A new process is proposed for the production Transport phenomena in very high rate and separation of the alkali metals at ambient electrolysis (up to 100 A/cm2) will be temperature. Alkali amalgam from comnercial quantitatively described to provide usable mercury-chlorine cells is transferred into an guidelines for the design and conduct of AICI3-PC electro-refining cell, where the technologically significant processes (e.g., alkali metal is dissolved anodically from the electromachining and electroforming) . An amalgam, and deposited in a pure metallic understanding of the role of anodic oxide form at the cathode. The lean amalgam is layers, particularly with reference to periodic returned to the chlorine cell where it is fluctuations of anode potential, will be again enriched with respect to the alkali sought, so as to allow rational selection of metal. electrolytes in electromachining. Development of new techniques for the electromachining of high strength materials will receive * increased attention. Abstract of LBL-1111, September 1972. c. Electrochemistry in Nonaqueous Solvents b. Reduction of Earth-Alkali Metals in Propylene carbonate The feasibility of development of techno Leonard Silvester and Charles W. Tobias logically significant processes for the recovery of reactive metals from their common Exploratory work has been initiated on the salts and for their separation and purification electroreduction of calcium at ambient tem using propylene carbonate as an ionizing perature by using propylene carbonate and solvent medium will be further explored. Work related aprotic solvents. will continue in three areas: 1) preparation of alkali metal alloys (such The long-term stability of the AICI3-PC as NaK) directly from halides, system as used for electroTeduction, separa 2) separation of a single alkali m;i.aA from tion, and electrorefining of Group I and II mixtures of alkali metals, metals is being investigated. 3) recovery of earth alkali metals. S0fi75 -42- Lepending or, funding levels, and on avail 2. Jacob Jorne", Electrochemical Behavior of ability of qualified personnel, work on the the Alkali Metals in Propylene Carbonate separation and recovery of rare earth will (Ph.D. thesis), LBL-1111, September 1972. also be undertaken. 3. K. IV. Beach, R. H. Muller, and C. IV. Tobias, Light Deflection Effects in the Interferometry of One-Dimensional Refractive Index Fields, 5. 1972 PUBLICATION'S AND REPORTS LBL-1102, September 1972. Charles IV. Tobias and Associates 4. Koaru Kojima and Charles IV. Tobias, Solution-Side Transport Processes in the Journals Electropolishing of Copper in Phosphoric Acid, LBL-1106, July 1972. 1. K. IV. Beach, R. H. Muller and C. IV. Tobias, A One-Dimensional Focusing Critical Angle 5. Koaru Kojima and Charles W. Tobias, Refractometer for Mass Transfer Studies, Rev. Interpretation of the Impedance Properties of Sci. Instr. 43, 925 (1972). the Anode-Surface Film in the Electropolishing of Phosphoric Acid, LBL-1129, August 1972. 2. L. F. Silvester, J. J. Kim, and P. A. Rock, Thermodynamics of Hydrogen-Isotope-Exchange Papers presented Reactions: I. Electrochemistry in Dimethylformamide, J. Chem. Phys. 56_, 1863- 1. Charles IV. Tobias, Some Fundamental Aspects 1869 (1972). of Industrial Electrochemical Processes, keynote address, 23 rd meeting of the LBL reports International Society of Electrochemistry, Stockholm, Sweden, August 27-September 2, 1972. 1. Koaru Kojima, A Study of the Mechanism of Electropolishing of Copper (Ph.D. thesis), 2. Charles W. Tobias, New Directions in LBL-140, February 1972. Electrochemical Engineering, Acheson Medal Address, 142 nd meeting of the Electrochemical Society, MiamTTteach, Florida, October 8-13, 1972. -43- Rolf H. Muller, Principal Investigator Presently, the electrode center is about 10 ym higher than the edges. 1. OPTICAL STUDIES OF INTERFACIAL PHENOMENA We now find that the observed interference a. Interferometry of Mass Transfer Boundary fringes agree with the predicted fringes to layers within a vertical displacement of 10 ym, even though some extraneous fringe shift and Frank R. McLarnon, Rolf H. Muller, and diffraction patterns are still present. Mew Charles W. Tobias polishing abrasives are new being tested as a means of further reducn.^ the rounding of When interferometric measurements of mass the electrode surface. This effort is transfer boundary layers are made, light expected to reduce the uncertainty in the deflection in the refractive index field position of the electrode surface, the gives rise to errors! JJI the resulting extraneous fringe shift and the diffraction inter ferograms. In order to check the com patterns in the interferograms still further. puted effect of these errors, we investigated the electrodeposition of Cu from a stagnant A function relating electrolyte concentra layer of 0.1 M C11SO4 at 10.0 mA/cm2. tion to a 4th order polynomial in vertical Preliminary results showed interference distance from the electrode surface has been fringes of the predicted shape, but they were found to successfully approximate the con displaced vertically (normal to the electrode centration profiles we have studied. This working surface) by 20-50 um from the pre functional relationship will be used in the dicted fringes. Also, an extraneous fringe interpretation of concentration boundary shift and diffraction fringes at the layers resulting from convective transport. electrode/electrolyte interface have been observed. Convective transport boundary layers in channel flow will now be investigated by Factors contributing to these undesirable interferometry and focusing refractometry phenomena have been found to be (a) small and interpreted using light deflection con errors (less than 0.5 mm) in focusing, (b) siderations and polynomial curve fitting. small errors (less than 0.2°) in the align ment of the electrode surface with respect to the incident light beam (specimen beam), (c) small deviations from flatness (less 1. IMRD Annual Report, LBL-425 (1971), p. 39. than 20pm) of the electrode surface. b. Automatic Ellipsometry The following corrective measures have been taken: (1) Eleven sets of scale lines H. Jorg Mathieu and Rolf H. Muller have been etched on one of the glass sidewalls at regular intervals along the 1.00 meter A self-compensating automatic ellipsometer electrode length to ensure reproducible with fast response and high resolution has focusing. (2) A new pair of electrodes, each been built for the measurement of thin films with one side face (3.81 cm high) polished on r.etal surfaces. flat to within 0.02° and optically smooth, has been built and installed. Thus, the The principal purpose of the construction incident light beam can be aligned within was to reduce the time (several minutes) 0.02° by reflection from the side faces. required for a manual measurement, so that (An uncertainty of 0.1° in the test beam transient films could be monitored contin angle may produce a 20 um uncertainty in the uously. A slewing rate of 1.8°/ms over a apparent location of the 10 mm wide electrode dynamic range of ±27° has been achieved for surface.) (3) Each electrode working surface the azimuth readings of polarizer and has been machined and polished to be at right analyzer. This rate compares to a value angle to the polished side face used for of 4°/s reported before for a partially alignment by reflection. A right-angle functioning instrument, based on the same polishing jig of a new design has been used operating principles. 1 Under conditions for this operation. Thus, the test beam which result in an azimuth change of a full can now be aligned parallel to the working turn for a film thickness increment of one electrode surface to approximately 0.05°. wavelength, we therefore expect to be able (4) New polishing techniques for the working to follow film growth at rates up to 2.5ym/s. surface of the electrode have been developed to reduce the rounding of the electrode edges. Another purpose of the construction was -44- PM T A AF 2 3 4 5 6 DVM Fig. 1. Block diagram of self-nulling ellipsometer A Analyzer AF Analyzer Faraday cell C Collimator CT Controller DVM Digital voltmeters (2) F Filter (X = 546.1 mp) G Galvanometer light beam oscillograph (channels 1-6) L High pressure mercury arc lamp M Modulation amplifier (10 kHz) 0 Oscillator P Polarizer PF PolarizeT Faraday cell PM Photo detector PS High Current Power Supply (dc) Q Compensator S Specimen T Telescope (XBL 7212-4927) to maintain an optimum resolution of the mechanical rotation of polarizer and analyzer electronic azimuth readings in the fast mode prisms, over the dynamic range. A control of operation and to increase it over the best signal is derived from a 10 kHz modulation manually obtainable resolution of 0.01° in a of the Faraday rotation to drive a null- slower mode. With a uniform, highly reflec seeking control loop. Current measurements ting surface, a resolution of 0.01° has been of the Faraday coils are used as azimuth demonstrated in the fast mode. This resolu readings and displayed in two digital volt tion decreases to 0.08° with a surface of meters, calibrated in degrees of azimuth poor optical quality. An increase in response rotation with readings to 0.001°. The signals time up to 5s (slow mode) has provided are also recorded by a multiple-channel light resolution down to 0.002'. Under the condi beam oscillograph. A schematic of the tions specified above, this resolution in instrument is given in Fig. 1. azimuth corresponds to a detection limit of approximately 0.01 monolayers of Kr and In situ observation of transient anodic proportionally less" for larger molecules. films on metal surfaces in a newly-built cell with exchangeable electrodes has been The instrument employs an electronic initiated. rotation of the plane of polarization of incident and reflected light by means of two Faraday cells, to take the place of the 1. H. P. Layer, Surf. Sci. 16, 177 (1969). -45- c. Computer Programs for Ellipsometry residual gas pressure of 10"" Torr. To en sure low reactivity with ambient gases and H. Jorg Mathieu to cover the widest possible range of optical phase changes due to reflection, Ag, Cr A short program (PSIDEL) has been written and Si are used as substrate materials. MgF2 to convert ellipsometer circle readings into and cryolite were chosen as the film relative phase A and amplitude parameter IJJ, materials. These materials are transparent The program determines the zone of a measure in the visible and are well characterized ment and provides two- or four-zone averages. optically and mechanically. To avoid con- tamination of the substrate, its deposition The correction of systematic errors of (on glass) in the vacuum chamber is followed analyzer, polarizer, and compensator circle immediately by the deposition of the thin reading, together with the deviation of the film. The substrate is radiantly heated from compensator from ideal retardation and a hot tungsten wire. Substrate and source transmission, has been incorporated in the temperatures are measured by Pt-Rh thermo computation of optical constants of bare couples . surfaces derived from measurements in one zone (program OZQM). White light interference colors for dif ferent angles of incidence are recorded Several previously used programs for the photographically using a positive, 35 mm interpretation of ellipsometer measurements color film. In addition, interference have been modified. Two- of them have been spectra are recorded photometrically using converted to Fortran IV language, which per a previously-built ratio-recording double mits the use of complex arithmetic and results beam spectrophotometer.1 Local thin film in shorter programs. Program CMER computes thickness is then derived (a) from the relative phase A and relative amplitude interference colors and (b) from the inter parameter ty for reflection from bare surfaces ference spectrum. from given values of the optical constants n and k. Program CMOC computes complex Results thus obtained are compared with optical constants from relative phase and step height determinations at the edge of amplitude or ellipsometer readings of the film by two other techniques: (1) d polarizer and analyzer circles in different mechanical stylus instrument and (2) inter zones. Two other programs have been modified ference microscopy. In addition, ellipsom to permit the use of ellipsometer circle etry will be used as an overall check. readings, rather than relative phase and amplitude parameters ns input and output variables. Program FILM computes the thick ness and complex refractive index of a single, 1. J. H. Turney, Profiles of Supermeniscus absorbing film on an absorbing substrate Electrolyte Films on Partially Submerged from ellipsometer readings in different Gas Electrodes (M. S. thesis), LBL-171, zones. Program DPLOT calculates and plots August 1971. ellipsometer readings for films of given optical constants and thickness. Programs for eliminating the effect of 2. RESEARCH PLANS FOR CALENDAR YEAR 1973 birefringent cell windows from ellipsometer measurements, together with the effect of Rolf H. Muller imperfections in polarizers and compensators, are being developed. Measuring techniques, now sufficiently developed for the interferometry of mass- d. Optical Interference in Thin Films transfer boundary layers, will be used for the detailed study of convective transport Charles R. Brown and Rolf H. Miiller in electrolysis. As a test of generalized interference The automatic ellipsometer will be used colors which have been computed previously,1 for the study of transient anodic surface the thickness of films derived from inter layers on metals. New computational procedures ference colors will be compared to thickness for the efficient interpretation of ellipsom determinations which do not depend on the eter measurements and the elimination of the optical properties of the film or substrate. effect of optical imperfections will be developed. Thin film-substrate combinations are prepared by vapor deposition on chemically Tests of the effect of phase change in and thermally cleaned glass slides in a optical reflection from a metal substrate. -46- on the derivation of film thickness from LBL reports interference colors, will be completed. 1. K. W. Beach, R. H. Muller, and The effect of microscopic surface roughness C. IV. Tobias, Light Deflection Effects in the on mass transfer with high flow rates in Interferometry of One-Dimensional Refractive narrow channels will be investigated (see Index Fields, LBL-1102, Sept. 1972 (submitted Paper 2a of Prof. Tobias' section). to J. Opt. Soc. of America). The propagation of electrical breakdown phenomena will be studied in the investiga 2. R. H. Muller, Double Beam Interferometry tion of anodic oscillations (see Paper 2b of for Electrochemical Studies, chapter in book Prof. Tobias' section). Advances in Electrochemistry and Electrochemical Engineering, Vol. 9 (to be published by lnterscience); LBL-11S Rev., March 1972. 3. 1972 PUBLICATIONS AND REPORTS Rolf H. Muller and Associates Journals and books 1. K. W. Beach, R. H. Muller, and C. IV. Tobias, A One-Dimensional Focusing Critical Angle Refractometer for Mass Transfer Studies, Rev. Sii. Instr. 43, 925 (1972)(LBL-153). -47- John Newmzn, Principal Investigator significantly nonuniform distribution may result. The degree of nonuniformity has 1. ANALYSIS OF POROUS ELECTRODES WITH been analyzed, and the problem of optimal SPARINGLY SOLUBLE REACTANTS location for the reference electrode has been discussed.1 John S. Dunning, Douglas N. Bennion, and John Newman The galvanostatic operation of flooded 1. J. Newman and J. E. Harrar, Potential porous electrodes employing metal/metal salt Distribution in Axisymmetric Mercury-Pool couples was analyzed. A model was developed Electrolysis Cells at the Limiting Current, for a single circular pore configuration LBL-1155, September 1972. which accounts for the effects of differing equivalent volumes of the solid reactants. The model also includes effects of variation in solution properties and effects of com- 3. MASS TRANSFER TO A NON-NEWTONIAN FALLING plexing of the sparingly soluble salt with FILM the bulk electrolyte anion. Overpotential was computed as a function of time for the Charles Mohr and Michael C. Williams solid-film and solution-diffusion versions of the model. The solid-film model shows a The Ellis rheological model was used to linear overpotential-time relationship and describe the hydrodynamics of a falling non- nearly uniform current distribution. The Newtonian liquid film. A perturbation anal solution-diffusion model shows a variety of ysis for the two resulting parameters in the overpotential-time curves, based on different convective diffusion equation for diffusion physical parameters, and may generate ex of a gas into the film resulted in a tremely nonuniform current distributions, hierarchy of equations which were readily often with maxima inside the pore. These solvable via computer. Results indicate persistant nonuniform current distributions that this is an accurate and economical way are important in terms of the analysis of to allow for variation of system and fluid cycling behavior. parameters in those situations requiring precise knowledge of the effect of the non- parabolic velocity profile. 2. POTENTIAL DISTRIBUTION IN AXISYMMETRIC MERCURY-POOL ELECTROLYSIS CELLS AT LIMITING CURREOT 4. MASS TRANSFER IN CYLINDRICAL COUETTE John Newman and Jackson E. Harrar FLOW Uniformity of potential at the working Charles Mohr and John Newman electrode has become an important criterion in the design of electrolysis and coulometry Theoretical equations for mass transfer cells. Deviations of the actual electrode to a section of the inner electrode for potential from the desired control potential laminar cylindrical couette flow weie derived may cause significant errors in coulometry and analytical solutions obtained. The and unwanted product distributions in electro- effects of curvature, the nonlinear nature organic syntheses. of the velocity profile, and elliptic effects at the ends of the diffusion layer were Because excursions of the working electrode included to extend the results of a previous potential are minimized, the preferred con treatment.* The mass transfeT Nusselt number figuration of a mercury-pool type cell is the was found to be a function of the Peclet axisymmetric arrangement in which the counter number and the ratio of electrode length to electrode separator is placed concentrically inner cylinder radius.2 Corrections due to above the pool. When the separator is elliptic effects dominated for short parallel to the pool, and of neaTly the same electrodes, but as electrode length approached diameter, a virtually uniform potential dis the inner cylinder radius, curvature and tribution should be achieved, and has been velocity profile corrections became pre experimentally realized. For cells in which dominant. For typical liquid systems of the separator is considerably smaller in interest, these corrections totaled less than diameter than the pool, however, a 10S. -48- the surface of a rotating disk as a boundary 1. D. R. Gabe and D. J. Robinson, condition for the convective diffusion Electrochimica Acta, JJ, 1121 (1972). equation, the time-dependent concentration 2. C. Mohr and J. Newman, LBL-1141, September distribution in radial and normal direction 1972. has been solved numerically. This solution agrees well with asymptotic solution for both close to and far away from the concentration jump. 5. CORROSION OF AN IRON ROTATING DISK The solution far away from the concentra Nader Vahdat and John Newman tion jump agrees with values obtained for the radially independent solution. The corrosion rate of iron rotating disks was predicted by calculating the current and potential distributions on the surface of the disk.1'' The results show that the sur 8. A COMPOSITE EXPANSION FOR THE WAKE JUST face towards the center of the disk is BEHIND A FLAT PLATE corroded at tb» maximum rate, while the edge of the disk remains almost unattacked, due John Newman and D. Theodore Scalise to passivation, in agreement with experimental observations. The effect of different param The fluid flow in the boundary layer adjoin eters on the corrosion behavior was also ing all solid surfaces presents a singular- considered. perturbation problem. A change in the bound ary conditions may generate additional Tegions in which different treatments are necessary. 1. N. Vahdat and J. Newman, LBL-896, July 1972. This singular-perturbation property is exhibited in Goldstein's1 classical investiga 2. N. Vahdat, M. S. thesis, LBL-873, June tion of a fundamental problem of fluid 1972. mechanics--that of determining the fluid velocity distribution in the near-wake boundary layer of a flat plate. Using one set of coordinate variables in the series 6. MASS TRANSFER COEFFICIENT IN POROUS FLOW expansion, he found an approximate solution THROUGH ELECTRODES valid only for the inner region near the plane of the plate; with another set of Harry H. Yip, John Newman, and Peter Appel coordinate variables he found an approximate solution valid only for the outer region, The flow system used previously has been the part of the boundary layer lying farther modified extensively. The electrolyte from the plane of the plate. solution now is pumped continuously from a reservoir through a microfilter (0.2u pores) 2 to an over flow tank, from where it flows back In this contribution, we construct a to the reservoir. uniformly valid expansion to the classical near-wake problem, using the method of matched asymptotic expansions. This is com The flow system is contained in a constant- pared with Goldstein's inner and outer solu temperature hood to minimize natural con tions at a downstream distance of half the vection effects. plate length. The great improvement is evident in the fact that whereas the old solu Flow disturbances at the entrance and exit tions are discontinuous in the central region, of the packed bed have been diminished by the new solution is continuous throughout containing the packing material between two the whole flow domain, merging to the old fritted glass disks. Also, the current col solutions at both extremities. lector—previously a stainless steel plate with holes drilled in it--has been replaced with a stainless steel screen. 1. S. Goldstein, Concerning Some Solutions of the Boundary LayeT Equations in Hydro 7. THE RADIALLY DEPENDENT CONVECTIVE WARBURG dynamics, Proc. Cambridge Phil. Soc. 26, PROBLEM 1-30 (1930). 2. John Newman and D. Theodore Scalise, An John Newman and Peter Appel Improved Solution to the Classical Near-Wake Boundary -Layer Problem, March 1973, LBL- Using an alternating concentration jump on 1430. -49- 9. MASS TRANSFER TO A PLANE BELOW A ROTATING The results allow the determination of time DISK AT HIGH SCHMIDT NUMBERS constants characteristic of decay due to an electrode reaction and due to a redistribution Robert V, Homsy and John Newman of charge within the double layer during the transient process. The Lighthill transformation is used to determine-1 the rate of mass transfer at high Schmidt numbers to a disk imbedded in a stationary plane below a rotating disk. This 1. John Newman, Ohmic Potential Measured by diffusion-layer solution breaks down near Interrupter Techniques, J. Electrochem. Soc. the center and leading edge of the mass- 117, 507-508 (1970). transfer disk. Solutions in these regions 2. John Newman, Resistance for Flow of Cur have been obtained previously2*3 g^d are rent to a Disk, ibid., 113, 501-502 (1966). applied to this problem. A uniformly valid 3. Kemal Nisanciojlu and John Newman, The composite solution for the local Nusselt Transient Response of a Disk Electrode, number is determined along with an expression LBL-1109, July 1972. for the average mass-transfer rate to the 4. Kemal Nisancioglu and John Newman, The disk. Transient Response of a Disk Electrode with Controlled Potential, LBL-1173, October 1972. An electrochemical experiment has been designed to measure the rate of mass transfer to the plane, using a copper deposition reaction at limiting current. 11. CURRENT DISTRIBUTION AT A ROTATING SPHERE BELOW THE LIMITING CURRENT Kemal Nisancioglu, Robert V. Homsy, and 1. Robert V. Homsy and John Newman, Mass John Newman Transfer to a Plane Below a Rotating Disk at High Schmidt Numbers, LBL-1420, January 1973. The rotating spherical electrode finds use 2. John Newman, Mass Transfer to the Rear of in electrochemistry because it exhibits a a Cylinder at Hif^i Schmidt Numbers, Ind. Eng. uniform primary-current distribution and thus Chem. Fund. £, 553-557 (1969). is applied in current-step niethods1 and high- 3. John Newman, The Fundamental Principles rate dissolution studies.- The limiting- of Current Distribution and Mass Transport in current distribution for this geometry has Electrochemical Cells, in Electroanalytical recently been calculated,2'3 and the distri Chemistry, Vol. 6, edited by A. J. Bard bution below the limiting current is also of (Marcel Dekker, New York, 1973), pp. 187-352. interest. The numerical method, which has been applied effectively to a number of electrode 10. THE TRANSIENT RESPONSE OF A DISK geometries in the pastf"? is adapted and ELECTRODE successfully tested for the rotating sphere. A number of case problems need to be worked Kemal Nisancioglu and John Newman out on the computer before reporting the results. The current-interruption technique is a commonly used method for the assessment of the ohmic drop in the solution in electro- analytical applications. One question arises 1. E. Mattson and J. O'M. Bockris, as regards the physical significance of the Galvanostatic Studies of the Kinetics of objrdc drop measured by such techniques. 1 Deposition and Dissolution in the Copper + Another question is more particular to the Copper Sulphate System, Trans. Faraday Soc. disk electrode, which exhibits a nonuniform 55_, 1586-1601 (1959). current distribution,2 and the manner by 2. Der-Tau Chin, Convective Diffusion on a which the charge decays after the interruption Rotating Spherical Electrode, J. Electrochem. at the double layer poses a nonelementary Soc. 118, 1434-1438 (1971). problem. 3. John Newman, Mass Transfer to a Rotating Sphere at High Schmidt Numbers, ibid.. 119, A mathematical model is proposed to treat 69-71 (1972). the transient response of a disk electrode to 4. John Newman, Current Distribution on a changes in the applied current3 and applied Rotating Disk below the limiting Current, potential.4 Analyses reduce to well-defined ibid., 113, 1235-1241 (1966). boundary-value problems, which yield solutions 5. John Newman, The Diffusion Layer on a in terms of newly defined eigenfunctions. Rotating Disk Electrode, ibid., 114, 239 (1967). -50- 6. W. R. Parrish arj John Newman, Current 2. W. Tiedemann and J. Newman, Growing Distribution on a Plane Electrode below the Porous Layers, J. Electrochem. Soc. 119, Limiting Current, ibid 116, 169-172 (1969). 186-188 (1972). 7. W. R. Parrish and John Newman, Current Distributions on Plane, Parallel Electrodes 3. W. Smyrl and J. Newman, Detection of in Channel Flow, ibid., 117, 44-48 (1970). Nonuniform Current Distribution on a Disk Electrode, J. Electrochem. Soc. 119, 208-212 (1972) (UCRL-20380). 12. RESEARCH PLANS FOR CALENDAR YEAR 1973 4. W. Smyrl and J. Newman, Ring Disk and Sectioned Disk Electrodes, J. Electrochem. John Newman Soc. 119, 212-219 (1972). The mass-transfer coefficient for packed 5. J. Newman, Book Review: Ionic Interac beds at low Reynolds numbers will be measured tions edited by S. Petrucci, Science, 175, by the electrochemical reduction of ferri- 1351^1452 (1972). cyanide ion. The local mass-transfer coef ficient for turbulent flow on a rotating disk 6. D. Bennion and J. Newman, Electrochemical will be studied by electrochemical deposition Removal of Copper Ions from very Dilute of copper. Calculations should reveal the Solutions, J. Appl. Electrochem. 2, 113-122 effect of ionic migration on limiting cur (1972). rents in turbulent-flow situations. Lead will be removed from a PbSOj saturated Paper presented sulfuric acid solution by electrochemical deposition in a flow-through porous electrode. 1. J. Newman. Corrosion of Iron Rotating Disks, lecture at Sandia Laboratories, Experiments are contei^Jlated for measuring Albuquerque, New Mexico, July 24, 1972. the time constant for transient current or potential changes for a disk electrode in LBL reports the absence of concentration variations. A related effort involves the calculation of 1. J. Newman and W. Smyrl, Fluid Flow in a the alternating-current impedance for a disk Propagating Crack, LBL-824, April 1972. electrode with consideration of concentration variations and the nonuniform ohmic potential 2. N. Vahdat, The Corrosion of Iron Rotating drop to the disk. A rotating disk can also Disks, M. S. thesis, LBL-R73, June 1972. establish a convective flow pattern on the bottom of the cell. The mass transfer to 3. N. Vahdat and J. Newman, Corrosion of an this plane will be studied by the electro- Iron Rotating dsk, LBL-896, July 1972. deposition of copper. 4. W. Tiedemann, J. Newman, and D. Bennion, The analysis of a rotating ring-disk The Error in Measurements of Electrode system with nonuniform current densities shows Kinetics Caused by Nonuniform Ohmic Potential promise of a successful and fruitful com Drop to a Disk Electrode, LBL-890, June 1972. pletion. Thermodynamic data for binary systems will be used to predict the thermo 5. K. Nisancioglu and J. Newman, The dynamic properties of ternary solutions. Transient Response of a Disk Electrode, Expressions for calculating virial coef LBL-1109, July 1972. ficients at low temperatures will be worked out, using Lennard-Jones type interaction 6. C. Mbhr, Jr. and J. Newman, Mass Transfer potentials. Studies may be begun on electro in Cylindrical Couette Flow, LBL-1141, phoresis and electro-osmosis. September 1972. 7. J. Newman and T. Chapman, Restricted Diffusion in Binary Solutions, LBL-1142, 13. 1972 PUBLICATIONS AND REPORTS September 1972. John Newman and Associates 8. J. Newman and J. E. Harrar, Potential Distribution in Axisymmetric Mercury-Pool Journals Electrolysis Cells at the Limiting Current, LBL-1155, September 1972. 1. J. Newman, Mass Transfer to a Rotating Sphere at High Schmidt Numbers, J. 9. K. Nis,ancioglu and J. Newman, The Electrochem. Soc. 119, 69-71 (1972) Transient Response of a Disk Electrode wich (UCRL-20599). Controlled Potential, LBL-1173, October 1972. -51- 10. R. Homsy and J. .Newman, Mass Transfer to Translations a Plane below a Rotating Disk at High Schmidt Numbers, LBL-1420, December 1972. 1. J. Besenhard and H. P. Fritz, Investigation of the Behavior of Graphite Electrodes in 11. K. Ni^ancioglu and J. Newman, Diffusion Nonaqueous Electrolytes, Z. Naturforsch. 26B, in Aqueous Nitric Acid Solutions, LBL-1428, 1225-1229 (1971). Translated bv R. Homsy, December 1972. UCRL-Trans.-1476, April 1972. 12. D. Scalise and J. Newman, An Improved 2. B. Kabanow and A. Frumlcin, On the size of Solution to the Classical Near-Wake Boundary Electrolytically Generated Gas Bubbles, Z. Layer Problem, LBL-1430, December 1972. Physik. Chem. 165A, 433-452 (1933) and 166A, 316-317 (1933)': Translated by J. Newman, UCRL-Trans-1485, September 1972. 3. E. Levart and D. Schuhmann, Cn the Dif- fusional Impedance for a Rotating Disk Electrode in the Range of Very Low Frequencies, Coll. Czech. Chem. Comm, 36, 866-876 fl971). Translated by C. Mohr, UC'RT-Trans-1484, August 1972. -52- D. PHYSICAL CHEMISTRY Leo Brewer, Principal Investigator 5. SOLUBILITIES OF INTERMETALLIC COMPOUNDS IN IRON 1. THERMODYNAMIC PROPERTIES OF THIRD AND FOURTH GROUP TRANSITION METALS WITH PLATINUM Leo Brewer GROUP METALS Work is progressing on the compilation John P. Stiga and Garry B. Bullard of data on solubility of intermetallic precipitates in iron and the prediction of Vaporization studies are in progress to behavior of all the metallic elements when establish the dependence of Gibbs energy dissolved in iron. upon concentration in solutions of strongly interacting metals. It is planned also to initiate experiments with high temperature cells, using solid electrolytes. 6. THERMODYNAMIC PROPERTIES OF Mo AND ITS COMPOUNDS Leo Brewer 2. APPLICATION OF THE REGULAR SOLUTION THEORY TO METALLIC SOLUTIONS Work is proceeding on this compilation of experimental data and predicted values. John P. Stiga A modification of the regular solution theory that uses promotion energies to obtain 7. THERMODYNAMIC PROPERTIES AND PHASE cohesive energies is being tested for metallic DIAGRAMS OF THE ALLOY SYSTEMS OF THE solutions that do not have strong interactions LANTHANIDE AND ACTINIDE METALS due to electron transfers. Leo Brewer The calculations of the phase diagrams of 3. OPTICAL PUMPING IN RARE GAS MATRICES the alloy systems of the lanthanides and actinides are progressing and should be com Dan D. Jensen pleted by the end of 1973. In previous work* on the spectrum of Mg in rare gas matrices, it was not possible to identify some features, and it was sug 8. RESEARCH PLANS FOR CALENDAR YEAR 1973 gested that they might be due to absorption from optically pumped levels of Mg. Leo Brewer Absorption experiments with and without the wavelength required for the proposed optical The studies of vaporization of transition pumping gave the same absorption features. metal alloys will be continued and be Thus they could not be due to optical punping supplemented by high temperature electromotive and are probably Mg2 features. force measurements. lferk on thermodynamic and phase diagram 1. L. Brewer and J. Wang, J. Mol. Spectr. compilations and calculations will be con 40, 95-102 (1971). tinued for some ternary phase diagrams, systems of molybdenum and iron with all elements, and systems of the lanthanides and actinides with all elements. 4. COMPILATION OF TERNARY PHASE DIAGRAMS In addition, it is expected that manu Shih-Ger Chang and John Chipman scripts on the low-lying electronic states of CaO and the transition probability of the Preparation of updated ternary phase Swan bands of C2 will be completed. If diagrams for the Metals Handbook of the support can be obtained, it is planned to American Society for Metals has been largely do additional work on the matrix isolation completed and should be finished early in of oxides of Mg and Ca. 1973. -53- 9. 1972 PUBLICATIONS AND REPORTS 7. Leo Brewer and J. Ling-Fai Wang, Ground State of Gaseous CaO; a Study of the Matrix Leo Brewer and Associates Spectra of Ca and CaO, J. Chem. Phys. 56, 4305-4309 (1972) (LBL-445). Journals and books 8. Leo Brewer, The Complexity of High Energy 1. Leo Brewer, Bibliography on the High Chemistry, Int. J. Sulfur Chem., Part B, Temperature Chemistry and Physics of Materials, Vol. 7, No. 1, pp. 31-35 (1972)(LBL-132). Part II. A. Spectroscopy of Interest to High Temperature Chemistry; B. Reactions 9. K. Wieland, J. B. Tellinghuisen, and Between Gases and Condensed Phases. Edited A. Nobs, The Band Systems E + X (2530-2740A) by M. G. Hocking, Vol. 16, Nos. 1, 2, 3, 4 of 127l2 and 129i2, and the Corresponding (1972J. System E ~ B of Br2 and CI 2, J. Mol. Spectr. 41_, 69-83 (1972) (LBL-860). 2. Leo Brewer, Relevancy of Research, J. Electrocheni. Soc. 119, 7-12C (1972) 10. D. D. Jensen, The Absence of Optical (LBL-195). Pumping in Matrix-Isolated Magnesium. J. Mol. Spectr. 43, 472-473 (1972) (LBL-803). 3. Leo Brewer and J. L. Wang, Infrared Absorption Spectra of Isotopic Ozone Isolated LBL reports in Rare-Gas Matrices, J. Chem. Phys. 56, 756-762 (1972) (LBL-141 Rev.). 1. Chin-An Chang, Temperature Dependence of Matrix Shifts, LBL-434 Rev., Feb. 1972. 4. Leo Brewer and C. Chang, Matrix Isolation of Pb and Pb2, J. Chem. Phys. 56, 1728-1740 2. Shih-Ger Chang, High Temperature (1972) (LBL-149). — Spectroscopy and Thermodynamics of Titanium Oxide and Vanadium Oxide (Ph.D. thesis), 5. Leo Brewer and C.-A. Chang, Electrical LBL-450, March 1972. Conductivities of Pb/Xe and Pb/SF6 Films, J. Phys. Chem. Solids 33, 457-459 (1972) 3. Leo Brewer and Paul R. Wengert, Transition (UCRL-20581). ~ Metal Alloys of Extraordinary Stability; an Example of Generalized Lewis-Acid-Base Interactions in Metallic Systems, LBL-807, 6. Leo Brewer and Joel Tellinghuisen, March 1972 (submitted to Metallurgical Quantum Yield for Unimolecular Dissociation Trans.). of 12 in Visible Absorption, J. Chem. Phys. 56, 3929-3938 (1972) (LBL-400). 4. Leo Brewer, Transition Metal Alloys—A Chemist's View, 18th Annual Conference on Magnetism and Magnetic Materials, Denver, Col. Nov. 28, 1972, (To be published in American Institute of Physics Conference Proceeding Series.) LBL-1424, Nov. 1972. -54- Charlea B. Harris, Principal Investigator 2. COHERENT ENERGY MIGRATION IN MOLECULAR SOLIDS 1. EXCITED TRIPLET STATES IN MOLECULES AND MOLECULAR CRYSTALS Charles B. Harris Charles B. Harris In order to describe the dynamics of exciton migration in the Fienkel limit, the Using new techniques recently developed in electronic states, the phonon states, and these laboratories which utilize simultaneous phonon-exciton coupling must all be explicitly optical and microwave pumping (Phosphorescence- considered in terms of the crystal states. Microwave-Double-Resonance Spectroscopy), we Indeed, it is the latter interaction that are studying molecular properties of excited determines the primary mechanism responsible triplet states and the dynamics of excited for electronic energy transfer in solids at state energy transfer (excitons) in molecular both high and low temperatures. At low tempera crystals. In oversimplified terms, ture the density of phonon states becomes luminescence from excited triplet states of sufficiently small that scattering between "isolated" molecules or mobile electronic the exciton wave vector states k by the phonons states, excitons, is monitored while a micro is expected to be much less frequent than wave field changes the orientation of the energy migration from lattice-site to lattice- electron spin in the triplet state. Ihis site. In the limit that the time between causes a change in the intensity or polariza scattering events approaches the radiative tion of the emission and forms the basis for or radiationless lifetime of the excited a technique capable of detecting as few as electronic state, a Frenkel exciton can be 10* molecules in excited states. Many features thought of as localized excitation propagating of triplet states can be studied by analyzing coherently as a wave packet at a velocity the microwave transitions and the triplet characteristic of both its energy and a linear phosphorescence. These include electron combination of the crystal k states. This distributions, nuclear-electron interactions, velocity is termed the group velocity and is excited state orbital symmetries, molecular given by distortions in excited states and, finally, the rates, routes, and mechanisms of inter- VgM = C2ir/h)(de/dk) (1) and intramolecular energy transfer. In addition we are investigating relatively and for a one dimensional crystal, z(k) unexplored but very important aspects of excited states involving molecular processes e(k) = E° + 28 cos ka (2) that occur on a time scale of 10"9 to 10"12 seconds. These are all fundamental in nature is the band dispersion associated with trans- and include such things as the primary lational equivalent interactions along a mechanism in energy migration, radiationless direction a. E° is the electronic energy relaxation phenomena, and electron and/or of the localized molecular excited state proton transfer between molecules. A variety while B is the effective intermoleculaT of mechanisms have been postulated and theoret interaction in the nearest neighbor approxi ically developed for these processes; however, mation. The distance £(k), which an exciton because of the extraordinary short time scale propagates in a coherent fashion without of the processes, few have been studied. Using changing either its direction or velocity Picosecond (lO"1^) Laser Spectroscopy certain is given by the lifetime of the coherent state, key aspects of the above phenomena are being t(k), times the group velocity of the wave- experimentally investigated in our laboratory. packet: In effect, a picosecond laser pulse is used to generate molecules in excited 'tates and *00 = V 00 • T(TO C3) then it is used to follow spectra.-copically g the time development of the systeir. under SL(X) is thus equivalent to a mean-free path investigation. Since it is possible to obtain and x(k) corresponds to a coherence time the complete spectra of molecules in for the wave vector state k. From a dynamical ~ 3 x 10"13 second using picosecond techniques, point of view the important feature of the primary events and time dependent changes coherent migration is that excitons can associated with excited states, such as energy propagate in the crystal at a variety of transfer, radiationless transitions, and the velocities and a variety of distances coherence associated with oscillating electric depending upon the k states populated. For dipolar states in molecules can be systemati the case under consideration the group cally investigated. velocity is zero at the top and bottom of the -55- band (k = 0 and ±ir/a) but SirS/h at the center individual k states acquire a width r(k), of the band (k = ±ir/2a). The extent to which given by the reciprocal of the coherence these states contribute to the propagation lifetime of the individual k states. r(k) of electronic energy in a given situation is is given by determined explicitly by the population dis tribution in the band. For a thermal distri bution, the numher of excitons N(k) propagating r(k) = [TOO]"1 = £ (t^,)'1 , C6) with a velocity Vg(k) is proportional to the density of states p(e) times the Boltzmann factors: where TI&' is the probability of an exciton initially in the kth state scattering via N(k) « P(E) • exp[-E(k)/kT] . (4) phonon-exciton interactions to a final state k'. The importance of coherent exciton migration is perhaps best illustrated by comparing it Thus, a proper description of the dynamics to the high temperature diffusion limited of exciton migration must include in addition random walk process. High temperature is to the stationary states of the crystal (a) characterized by a density of populated phonon the group velocities of excitons, (b) the or localized vibrational states sufficient to population distribution over the k states of limit the coherence lifetime of an exciton k the band, and (c~) the coherence times for the state by scattering it to other states in individual k states and hence an explicit the band. When this causes a change of k at model for phonon-exciton scattering. This a rate comparable to intermolecular energy description views the exciton as executing exchange, the delocalized wave vectors of a random walk migration in a time on the order the crystal are no longer good wavefunctions of the coherence lifetime and allows for long- for times exceeding (B)"1 and the proper range propagation via coherent migration in description of the exciton becomes the Fourier between scattering events. transform function of the exciton k states. In this limit the exciton, on the average, From an experimental point of view, this changes its velocity and direction at every model requires that careful attention be given lattice site and propagates in a diffusion to the relationship between the correlation limited random walk manner. For one-dimension time associated with exciton migration and al propagation the velocity of migration of the time scale of the particular experimental the exciton is simply an RMS average velocity, approach being employed. If, for example, and is proportional to CB) ' . Consider, for the experimental correlation time, which is example, a molecular crystal with nearest on the order of the reciprocal of the radiation neighbor molecules having an effective inter- field, is much shorter than T(k) (as is the molecular interaction of 1 cm" I and a lattice case for optical absorption), only manifesta spacing of 4 A The diffusion limited exciton tions of the coherent model are apparent from velocity is 10"2 cm/sec while coherent states the data. Similarly, when the experimental have a group velocity at the center of the correlation time is longer than r(k) for all band of l(p cm/sec. The mean-free path for k, only the random walk processes are displayed. diffusion is 4 A; however, for coherent states A reliable measure of phenomena connecting at k = ±ir/2a it is coherent migration and diffusion limited migration., such as phonon-exciton scattering, ft(k = ±it/2a) = (8n?l/t0 • x(k = ±ir/2a) . (5) Vg(k) and £(k), can only be determined when the experimental correlation time is on the It is self-evident that the coherence length order of t(k). It is on this basis that the can approach macroscopic dimensions if phonon- following outlined experiments provide a exciton scattering is weak [i.e., r(k) is direct probe into the dynamics of triplet long] and the excited states are long-lived excitons. (e.g., triplet states). In practice this is only achieved at very low temperatures where the distribution of phonon states approaches the T-+0 limit. At intermediate temperatures 3. ELECTRON SPIN RESONANCE ABSORPTION IN the principal limitation of r(k) is phonon- COHERENT TRIPLET EXCITON STATES exciton scattering. In such cases an exciton initially in a k state scatters to otheT k' Charles B. Harris states via phonon interactions in a time short compared to the radiative or radiation- An electron spin resonance theory for ex- less lifetime, but in a time long compared itons in a one-dimensional crystal composed to interroolecular exchange. As a result the of translationally equivalent molecules has coherence time is shortened, the mean-free been developed. The excited triplet band is path or coherence length is reduced, and the derived from intermolecular exchange coupling -56- between the ground singlet and excited triplet per unit time for an exciton being scattered states and is restricted to nearest neighbor by a phonon from k to k1, is displayed as a interactions. Furthermore, a Boltzmann "golden rule' rate distribution over the band was assumed. In zero magnetic field the triplet band is split ) exp[-EA(q)/kT] pfq) into three parallel bands by electron spin l/ifvk . " P,0") ]Cp„Cq' dipolar repulsions. The spin sublevel bands *7q7q' are designated as TX, ty, and tz and cor respond to states in which the electron spins are correlated in the x, y, and z molecular * I < kq I K^p |k'q' > I2 •SOq.k'+q') • CH) axes respectively. Electron spin transitions between any two spin sublevels are independent of k in zeroth order since the spin Hamiltonian Pe(k')> Pp(q')» and 0p(q) are the exciton is k independent. In first order, however, (subscript e) or phonon (subscript p) density selective spin-oTbit coupling between excited of states functions evaluated at the energy singlet molecular states and individual molec of the wave vector k', q', or q respectively. ular spin sublevels of the triplet results in X is the index which runs over the phonon a k-dependent contribution to the zero-field branches and EA(q) is the energy of the qth band splitting and hence a k-dependent elec wave vector of the phonon branch. The sum tron spin band-to-band transition between over phonon states q and q' is restricted to magnetic sublevels. The k dependence of the collisions that conserve both the total Larraor frequency, u>§, is given by energy and momentum of the initial and final exciton-phonon states < kq |. Wg_ :i s the exciton-phonon coupling Hamiltonian. In weak u£ = C'[2B cos ka] , (73 oscillating rf fields and steady state (dGjj/dt = 0) the electron spin resonance line where V is a factor related to the details shape function, g(w), is given by: of spin-orbit coupling. The proper representation for electron gM In. T,G. (12) k spin resonance absorption in terms of the k Bloch formalism is a set of Bloch equations, one for each k state in the band, whose The function g(u) was easily solved in two frequency components u$ are coupled by limits. The first, termed the strong scat phonon-exciton scattering. In the rotating tering case, occurs when (mo -too* )t%k' <<: 1 frame the presence of a weak oscillating rf and yields a homogeneously narrowed line field of the form centered at k = iir^a at high temperatures and corresponds ±0 the random walk limit. The second, when (IDO o>o )Tkk' >:> 1> corresponds JC(t) YH-:, • S COS (lit (8) to the coherent limit in which phonon-exciton scattering causes a change in the exciton connecting^Tx with TZ via the electron spin 1 operator S results in an in-phase, u, and states on a time (rkk ) slow compared to the out-of-phase, v, component of a complex moment, differences in the Larmor frequencies G, given by (ioc(*-oio* ). In such cases the individual k states of the triplet band can be sampled by the rf field. Assuming phonon-exciton scat G = u + C9) k k tering to be uniform in k, i.e., TifKi are equal for nil k and k', the electron spin resonance Modified Bloch equations were constructed absorption g(w) is the sum of (2k + 1) from the complex moments of each k state that independent Lorentz lines each centered at include the effects of phonon-exciton scat Wo* and weighted by the number of excitons tering phenomenologically through the scat occupying the state k with group velocity tering rates (lAkk') • The (2k + 1) equations Vg(k). The width of each Lorentz has a contribution from both a finite coherence lifetime t(k) and the homogeneous line with parameter T2(k). When a thermal distribution dGj/dt + [1/T (10 - 1(^-0)]]^ = - iyHjNM 2 characterizes the triplet band •Tr/a exp[4B(l-cos kaVkTI !(»)= dk (13) where N(k) is given by Eq. (4). T2(k) is the [10 + A|J. COS ka]2 + <52 homogeneous spin-spin relaxation time for the kth exciton state. 1/T(&», or the probability -57- A^sT is related to spin orbit coupling, T is Initially the problem was considered for temperature, 6 is the half width at half an ensemble consisting of one-dimensional height of an individual k state electron spin exciton bands and traps of only one energy. transition and is a measure of T(k). T£e trap intensity is given by I trap (T) = K N K p trap trap. = trapNtotal trap CO. "here K£rap Some specific features of Eqs. (10) aiid (13) is the radiative rate constants, Ntran is the are that: (a) The Larmor frequencies are trap population, and Ntotai is the total directly related to band energy insofar as a excited state population of the system. Since prescribed u$ couples the spin sublevels of an an excitation must either be in the exciton exciton whose energy is e(k). (b) The intensity bands or in the traps, the trap population is of the transition is directly related to the equal to the total population times PtrapCO, density of states function times a Boltzmann the probability that an excitation is in a factor. The relative heights of the transition trap. The number of energy levels in a one- maxima at - k = 0 and ±ir/2a results from the dimensional exciton band is equal to the Boltzmann factor evaluated at the maxima in number of identical molecules in the linear the density of states function at the top and chain forming the exciton. These energy bottom of the band and thus gives a value levels are labeled by the quantum number K, for the band width, (c) The broadening and the energy as a function of k is given function 6 gives an estimate of the phonon- by E(k) = 2$ cos ka, where 6 is the nearest exciton scattering, and hence a measure of the neighbor intermolecular interaction matrix coherence length £(k) and insight into the element and a is the lattice spacing in the mechanism of phonon-exciton scattering. The direction of the interactions. Using this dependence of g(co) on Mgp via (r^t)"''- might, energy relationship, the partition function for exarple, allow linear and quadratic z for the system can be written as phonon-exciton coupling to be distinguished experimentally, (d) The overall width of A/kT n /na [A 2SC1 cos ka the g(w) is determined by the spin-orbit z = 1 • e" * 2 2e- - ' )VkT_ ( coupling parameter A^g-r, hence the selectivity k=rr/na of spin-orbit coupling to the triplet spin sublevels can be determined, (e) The micro The zero of energy is taken at the energy of wave field selectively perturbs the electron the trap. The trap depth A is the difference spin of excitons whose group velocity is Vg(k), in energy between the trap energy and the and therefore the k dependence of other proc energy of the k = 0 level of the exciton band. esses such as the dynamics of trapping, This is the depth which is measured spectro- exciton-exciton annihilation etc. can be scopically since only the k = 0 level of the potentially studied in the coherent limit. origin absorbs or emits light. The first Thus, electron spin resonance provides a term in z is the Boltzmann factor for the direct technique for measuring band structures, trap level, the second term is for the non- intramolecular exchange interactions, phonon- degenerate k = 0 level of the band. The exciton scattering, and other related summation is over the remainder of the exciton phenomena. states which are doubly degenerate. The trap probability Ptrap = Vz, and therefore the intensity of trap emission I « 1/z. 4. EXCITON-TRAP INTERACTIONS IN THE COHERENT If a sample of known trap concentration is AND INCOHERENT LIMITS prepared and the trap depth A is measured spectroscopically, then 8 can be fit to the Michael D. Fayer and Charles B. Harris temperature dependence of the trap emission. As depicted in Fig. 1, two cases arise At liquid He temperature, phosphorescent depending upon the sign of B. If 6 is negative emission from organic molecular crystals in the exciton band spans an energy range from general comes from low energy traps in the A, the k = 0 energy, to A + 46, k = tir/a. If lattice although in some molecules, such as 6 is positive the band is inverted and it 1,2,4,5-tetrachlorobenzene and spans an energy range from A to A-4B. Thus 1,4-dibromonaphthalene, exciton emission is the sign of B in addition to the magnitude of also observed. The intensity of trap emission S determines the temperature dependence at is highly temperature dependent in certain the trap intensity. The above description temperature regions. This temperature gives excellent agreement with experiment for dependence is found to be explicable in terms crystals of 1,2,4,5-tetrachlorobenzene which of a theoretical framework, which includes have one-dimensional excitons and traps of a the method of energy transport in the crystal single energy. In order for this description and the magnitude and sign of the inter- to work, the system must reach equilibrium molecular interactions which give rise to the within the 1 msec lifetimes of the excited exciton bands in the crystal. states. This implies that excitations must -58- E(K) = E°+2£cosKa the states. Thus, coherent migration is established as the principal mode of exciton transport at liquid helium temperatures. Negative 0 Positive & In addition to the above discussion, the Exciton Band Inverted Band model developed for the temperature dependence of trap emission also treats multidimensional K = ±ir/ET/aa •N exciton bands, systems in which there are traps of several different energies, the Exciton Band effect of spin lattice relaxation between the three triplet sublevels, and systems in which the excitons and traps have different life 40 times. The model applies equally well to K = ±ir/a= fluorescent and phosphorescent traps. A Trap Depth Trap Depth 5. PHOSPHORESCENCE MICROWAVE DOUBLE RESONANCE SPECTROSCOPY IN EXCITED TRIPLET STATES William G. Breiland, Michael D. FayeT, Charles B. Harris, Allan M. Nishimura, Rodney M. Panos and Horst Schuch The technique of optically detected magnetic resonance (ODMR) in phosphorescent triplet Trap Level Trap Level states in magnetic fields and zero field recently developed in these and other Fig. 1. Trap level and exciton band for sys laboratories has provided experimentalists tems containing regular and inverted exciton with a tool capable of elucidating many bands. (XBL 7210-5704) features of triplets that have been difficult to obtain via conventional optical spectroscopy or magnetic resonance techniques. Since phosphorescence from the lowest triplet state be able to migrate to many trap sites in less to the ground state singlet is generally than the lifetime. The traps in the 1,2,4,5- specific from the individual magnetic spin tetrachlorobenzene samples are of such low sublevels, an analysis of the phosphorescence concentration that it is possible to distin microwave double resonance fl*!DR) spectra can guish between the two possible modes of yield spin-orbit symmetries of the individual exciton transport, coherent migration or random spin sublevels and hence the orbital symmetry walk hopping. In the coherent model the of the excited triplet state while a quantita velocity of an exciton is given by tive analysis of the changes in phosphorescence induced by microwave field or changes in the 3E v - 1 .« (2) magnetization yield relative intersystem crossing rates and relative radiative rates associated with the sublevels. In addition The average velocity at a temperature T is to parameters associated directly with radiationless and radiative processes in the / . sin(ka)e"-(2 8 cos ka)/kT triplet state, OEMR provides a measure of the 2Sa k electron distribution in excited states from g ti -(2$ cos ka)/kT an analysis of the spin Hamiltonian which in ro zero field include the spin dipolar, the T electron-nuclear hyperfine, and the nuclear quadrupole interactions. At liquid helium temperatures < V„ > *= 2700 cmVsec for 1,2,4,5-tetracfilorobenzene These techniques were extended to a variety which is more than adequate to insure of aromatics. equilibrium. In the hopping model, the median velocity is givsi by a. 1,2,4,5-tetrabromobenzene /2 V = I lrWJ (4) "hop g t In 1,2,4,5-tetrabromobe^zene the results showed that the phosphorescence originates 6 3 Vhop is found to be a factor of 10 smaller from the undistorted Bju(Tm*) state. The than (Vg > and is too small to allow the sample sign and magnitude of the zero-field parameters to come to equilibrium within the lifetime of were correlated to the spin dipolar interac- -59- tions in the •*Biu state and the large spin- 6. BREAKDOWN OF THE BORN-OPPENHETMER orbit contributions to the zero-field split APPROXIMATION IN RADIATIVE TRANSITIONS ting. Finally the excited triplet state values for the '°Br and 8^Br nuclear quadrupole Charles B. Harris, William G. Breiland and constants were obtained from an analysis of John C. Brock the zero field ODMR. Using phosphorescence microwave double resonance (PMDR) techniques, radiative con b. 1,4-dibromonaphthalene tributions from the individual spin sublevels to various totally symmetric vibrations in Work was completed on 1,4-dibromonaphthalene the ground singlet manifold were determined (DBN) single and doped crystals at 1.4°K. The for the lowest excited triplet state of zero-field splitting parameters were measured 1,2,4,5 tetrachlorobenzene (TCB). Both for several trap emissions. The results of h2-TCB and d2"TCB doped 1* m/m in durene were decay measurements by microwave-induced investigated and studies of napthalene doped adiabatic inversion and conventional in TCB are in progress. phosphorescence decay experiments were used to assign the symmetries of the triplet magnetic The TCB data demonstrated an explicit sublevels and also used to consider several nuclear coordinate dependence of electronic phosphorescence mechanisms. Orbital symmetries tutrix elements in the transition moment for of the triplet state of DBN were related to phosphorescence to the ground state vibrational the zero field electron dipolar interactions manifold. Theoretical egressions for the and were used to consider the magnitude of experimental observations developed within the experimentally measured zero-field the framework of the Condon and Herzberg- splitting. Finally, results of spin alignment Teller approximations fail to explain the Measurements of several traps in DBN neat results, implying that these approximations crystal were understood in terms of perturba are not general enough to give a good tion L, phonon interaction. description of radiative transitions in aromatic molecules. Additional terms in the Herzberg-Teller expansion and the breakdown of the Bom-Oppenheimer approximation were c. 1,4-dichlorobenzene shown to be alternatives which can be used to explain the experimental observations. Phosphorescence microwave double resonance (PMDR) spectra of p-dichlorobenzene have been Further insight into the specific depen obtained monitoring low temperature (~1.3°K) dence of the electronic wave function on trap phosphorescence and used to make symmetry nuclear motion may be gained by analyzing the assignments of various vibronic bands present normal coordinate of the vibration in in the spectra. In addition, ratios of relationship to the positions of the nodal radiative rate constants of individual triplet planes in Che excited electronic state. The spin sublevels to several vibrational states, ratio might be relatively unaffected, for the intrinsic lifetimes of all three triplet example, when the nuclear motion is associated spin sublevels, and selective polarized with a direction that keeps the nuclei in a phosphorescence from individual spin sublevels nodal plane of the electronic wave function. in various vibronic transitions have been Examination of these relationships is measured. An analysis of these data along presently under investigation. with explicit considerations of the p-chlorine perturbation on the aromatic ring elucidated the relationships between the orbital sym metries and the sign and magnitude of the 7. COHERENT SPIN PHENOMENA IN MOLECULAR zero-field electron spin dipolar irteractions EXCITED STATES in the triplet state for the halobenzenes. On these bases an assignment of the orbital Henry C. Brenner, William G. Breiland, symmetry of the lowest excited triplet state Terence E. Gough, Charles B. Harris, Ronald 3 of p-dichlorobenzene was established as Biu. L. Schlupp, and Horst Schuch d. Hexachlorobenzene In phosphorescent triplet states, or, for that matter, in any state where the emission Preliminary PMDR data have shown that the from the individual spin states is unresolved excited triplet state in hexachlorobenzene optically, it was shown earlier in these doped in hexamethylbenzene exhibits many of laboratories that modulation of the light at the same features as 1,4-dichlorobenzene and the Larmor frequency by an applied microwave the tetrahalogenated benzenes. field is unobservable unless the electric -60- dipole transition from the individual spin states has been measured* point-by-point by levels is of the same polarization. In many observing the phosphorescence intensity after excited triplet states, however, the spin two u/2 microwave pulses, the first to prepare sublevels of molecules have different spin- the coherent state and later a second to probe orbit symmetries and thus direct detection the instantaneous value of the coherence in of the- Larmor term is usually forbidden. the x-y plane by returning it to the z axis. Likewise, spin echoes in excited states have In other words, unlike conventional magnetic been optically detected by a three-pulse resonance where components along x or y may sequence TT/2, ir, and TT/2, the first two pulses be observed, optically detected magnetic being used to prepare the echo and the last resonance requires that the time-dependent pulse being used to monitor point-by-point density matrix describing the spin ensemble the formation of the echo in the x-y plane be displayed through the electric dipole by monitoring the effective population restored transition moment. The result is that only to z and the resulting change in the components along z in the interaction represen phosphorescence intensity at the end of a tation are directly observable via a change in sequence. the rf-induced modulation of phosphorescence intensity. Since the coherence of the spin Shallow traps in tetrachlorobenzene havo ensemble in the interaction representation is been studied by these techniques, and a confined to the x-y plane, phasing or dephasing homogeneous T2 of 4.1 ysec observed over the phenomena do not normally affect the temperature range 1.4 •+ 4.2°K has been phosphorescence polarization or intensity. obtained. The value of 4.1 usee should be Components of the spin in the x-y plane compared with the inhomogeneous T2 of correspond simply to unchanging saturated ~250 nsec. The echo decays are nonexponential, populations of the two spin sublevels in the showing a beat frequency of approximately laboratory frame. 0.9 MHz, indicating the presence of a splitting due to the unresolved 35ci hyperfine inter- The ability to restore the enseirble from echos matched by the inhomogeneous broadening the x-y plane to the z axis is directly in the conventional C. W. spectrum. proportional to the extent of dephasing. If the spins are completely dephased in the x-y We have also successfully observed via plane no further change along z can be brought optical detection rotary electron spin echoes about by an additional microwave perturbution and rotary echo trains in molecular excited and hence the population of the spin sublevels triplet states utilizing a T-2T'-2T-2T' ••• in the laboratory frame remains constant. On pulse train, where primed T pulses are 180° the other hand, if the ensemble can ba in phase shifted from unprimed T pulses. part restored from the x-y plane the spin Successful echoes have been observed for proto sublevel populations in the laboratory frame and deutero 1,2,4,5-tetrachlorobenzene. This will take on new values. The net result is sequence is being used to examine the correla that a further change in the phosphorescence tion function for eneTgy migration in intensity can be affected by additional molecular crystals containing both traps and microwave perturbations. It is expected excitons. This is accomplished by varying T. therefore that many manifestations of spin For very short T then, diffusion (« T3) can coherence can be experimentally observed via be neglected in the decay of the echo ampli optically detected magnetic resonance tech tudes while for long x diffusion becomes niques. Furthermore the technique is capable proportionately more important. in principle of measuring the ensemble coherence of as few as 104 excited state spins. Work has also begun on long-range guest-to- guest triplet energy transfer in mixed molec An apparatus has been constructed to allow ular crystals using optically detected the optical detection of electron spin echoes magnetic resonance techniques at low tempera of excited states. In the construction of tures. One potentially favorable system is such apparatus, careful attention has been phenazine and anthracene doped into diphenyl paid to the design of the microwave system host crystals. It is known from published to provide good impedance matching throughout. work in other laboratories that the phenazine Impedance matching is important for both the triplet is populated almost exclusively through efficient generation of high intensity the top zero field level by selective inter- microwave fields and to minimize unwanted system crossing from the singlet manifold, and reflection which distorts the pulse shapes. that the spin alignment is preserved upon Currently clean SO nsec pulses of any micro transfer to anthracene, whose triplet lies wave phase having W\ ~ A gauss are obtainable several hundred wave numbers lower. The throughout the region 1.0 to 8.4 GHz. transfer proceeds either by direct exchange Electron spin free precession in excited In collaboration with A. Pines. -61- transfer from phenazine triplet to anthracene process of generating multiple-light-pulse ground triplet, or by means of diphenyl trains of controlled amplitude. In addition triplet excitons generated from singlet excited to studying relaxation phenomena of states phenazine. Using this or another system where exhibiting an oscillating macroscopic electric light emission is more intense, we are attempt dipole moment, the lasers are being used to ing to induce microwave transitions in the study coherent times in exciton migration and donor triplet and monitor the modulated to study energy trapping in defect and phosphorescence of the acceptor triplet. This impurity sites in molecular crystals. could answer the question of whether the transfer proceeds through the host triplet. Using the techniques for optical detection of triplet state spin echoes, recently developed 9. RESEARCH PLANS FOR CALENDAR YEAR 1973 in this laboratory, we are attempting to determine spin memory times upon energy Charles B. Harris transfer from donor to acceptor. Our major emphasis will be to explicitly Finally, electron spin locking in molecular determine the nature of processes that destroy excited triplet states has been observed via the coherent transport of electronic energy optical detection. It is expected that vir in molecular crystals. This will be approached tually any process that affects spin lattice from two types of experiments. The first will relaxation in the rotating frame will be be to observe, as a function of temperature amenable to study. This should allow the and applied magnetic field, the averaging development of many new double resonance effect of the exciton optically detected techniques involving other spins (electron magnetic resonance resulting from exciton- and nuclear) and other processes such as phonon collisions. The EPR line-shape theory acoustic phenomena. for coherent excitons developed in these laboratories is dependent upon the relative size of (UQ* - wo ) and 1/Tidj', where J^ is an ESR transition frequency associated with 8. COHERENT OPTICAL PHENOMENA IN MDLECULAR a particular value of k and I/T^I is the EXCITED STATES probability per unit time for scattering of an exciton in a state k to a state k'. In the Charles B. Harris, Allan M. Nishimura, and coherence limit, (uj^ - oicT ) > 1/THC' and Robert M. Shelby the observed line is a superposition of individual ESR lines at different frequencies The coherence associated with an electric determined by the values of k. In the strong dipole transition of an ensemble of excited scattering, random walk case, (wo* - afcK ) « states can in principle be studied by using 1/Tkk'f and a homogeneously narrowed line at a sequence of light pulses similar to the the average value of k is expected. microwave pulses developed for spin states in section 7. The importance of such studies In the coming months we intend to study in understanding radiative and radiationless phonon exciton scattering processes in applied relaxation phenomena from excited electronic magnetic fields. As mentioned above, the states is that a direct measure of the exciton spin Hamiltonian parameters are k ensemble phase can in principle be directly dependent. Thus we expect that ask - uk> will measured via its relationship to the total depend in general on the magnitude and direc population left in an excited state after the tion of an applied magnetic field. Hopefully application of a given light pulse sequence. one can proceed snoothly from the coherent The dephasing of the moment by various limit to the strong scattering limit by relaxation routes has been a topic of current varying the field. One can then obtain interest in echo studies of electron spin information about the k dependence of I/TJ^I states, but experimental work has been limited and test out various models for phonon exciton to slow processes in the analogous photon scattering. The magnetic field studies will echo and relaxation phenomena due to previously be complemented by experiments in which we unobtainable optical pulses shorter than vary the temperature, which affects the values several tens of nanoseconds. In an attempt of I/TI&I directly. Secondly, the observation to overcome these difficulties and study of coherence times in excitons and the scat extremely fast processes (10~12 sec), we have tering processes will be probed by utilizing constructed high power Nd:glass and ruby pisosecond laser techniques. lasers, both of which have unusual stability and have been successfully mode-locked. In Specifically, by optically exciting a addition, single pulse selection has been crystal into an exciton state in a time on obtained by using a spark-gap-triggered the order of 10"12 sec with a picosecond Pockels cell. We are presently in the laser pulse, an ensemble of excitons are -62- produced in the k = 0 state of the crystal. 2. M. J. Buckley and C. B. Harris, The Lowest If no exciton scattering occurs, one expects Triplet State of Substituted Benzenes: I. The the time-dependent radiative decay from the Optically Detected Magnetic Resonance of exciton state to be a simple exponential to Paradichlorobenzene, J. Chem. Phys. 56, 137 both electronic origin and a vibronic origin. (1972)(LBL-128). If, however, the prepared k = 0 states scatter to other k states then the radiative emission 5. M. J. Buckley, C. B. Harris, and to the electronic origin will not be the same R. M. Panos. The Lowest Triplet State of as to the vibronic origin. In particular, Substituted Benzenes: II. Phosphorescence the vibronic origin will not appear exponen Microwave Double Resonance Studies on the tial until an equilibrium Boltzmann 3TT7T* State of p-dichlorobenzene, J. Am. Chem. distribution is established in the exciton Soc. 94, 3692 (1972)(LBL-409). band (via phonon-exciton scattering), at which point the scattering into individual k states 4. A. H. Francis, C. B. Harris, and equals the scattering out of the k states. A. M. Nishimura, Properties of the It is expected that the nonexponential Phosphorescent State of 1,2,4,5- behavior at early times reflects the scattering Tetrabromobenzene as Determined by Optica] "iv to nonzero k states with a rate proportional Detected Magnetic Resonance, Chem. Phys. to the exciton density of states function Letters 14, 425 (1972)(LBL-811). times the phonon-exciton scattering interaction. The time dependence of exciton emission meas 5. A. H. Francis and C. B. Harris, On the ured via fluorescence is expected to yield, Lowest Triplet State of Substituted Benzenes: therefore, the coherence time (the average III. Optically Detected Magnetic Resonance lifetime of a k state) and thus a direct Studies on the 3mr* State of 1,2,4,5- measure of the excitons mean free path in the Tetrabroriubcnzene and 1,2,4,5 crystal. Tetrachlorobenzene, J. Chem. Phys. 57, 1050 (1972)(LBL-472). Finally, models for bimolecular exciton- exciton annihilation will be developed and 6. C. B. Harris, Magnetic Resonance Absorption tested in the coherent limit. Coherent exciton in Coherent Exciton States of Molecular migration should allow two triplets to come Crystals, Abstracts of the XVII th_ Congress close enough to each other to undergo triplet- Ampere, Turku, Finland, August 1972. triplet annihilation at a very accelerated rate. This results in the loss of the two 7. C. B. Harris, Coherent Migration of Triplet triplets and the production of a ground state Excitcns in Molecular Crystals, in Proc. of singlet and a highly excited singlet. The the 3rd International Symposium on the highly excited singlet can relax to the first Chemistry of the Organic Solid State, Glasgow, excited singlet state and may then fluoresce. Scotland, September 1972. Fluorescence of this kind is known as delayed fluorescence. By studying the details of Papers presented the coherent migration process and thereby determining as a function of temperature the 1. C. B. Harris, Coherent Migration of Triplet probability of two triplets encountering Excitons in Molecular Crystals, 27th Annual each other, a model for the temperature Symposium on Molecular Structure and Spectros dependence of delayed fluorescence can be copy, Columbus, Ohio, June 1972. developed from a statistical basis. 2. C. B. Harris, Magnetic Resonance Absorption in Coherent Exciton States of Molecular Crystals, XVII th Congress Ampere, Turku, Finland, August 1972. 10. 1972 PUBLICATIONS AND REPORTS 3. C. B. Harris, Coherent Migration of Triplet Excitons in Molecular Crystals, 3rd Charles B. Harris and Associates International Symposium on the Chemistry of the Organic Solid State, Glasgow, Scotland Journals September 1972. 1. C. B. Harris and R. J. Hoover, Optically 4. C. B. Harris and R. M. Panos, Optically Detected Adiabatic Inversion in Phosphorescent Detected Magnetic Resonance in Substituted Triplet States and the Measurement of Intra Benzenes, 1972 Pacific Conference on Chemistry molecular Energy Transfer Process, J. Chem. and Spectroscopy, San Francisco, Calif., Phys. S6_, 2199 0972) (LBL-118 Rev.). October 1972. -63- 5. M. D. Fayer and C. B. Harris, Trapping 2. W. G. Breiland and C. B. Harris, Breakdown of Coherent Triplet Excitons in Molecular of the Born-Oppenheimer Approximation in Crystals, 1972 Pacific Conference on Chemistry Radiative Transitions, LP'-1411, November 1972. and Spectroscopy, San Francisco, Calif., October 1972. 3. W. G. Breiland, C. B. Harris, and A. Pines, Optically Detected Electron Spin a Echos and Free Precession in Molecular Excited Presented by States, LBL-1412, December 1972. LBL reports 1. C. B. Harris, Magnetic Resonance Absorption in Coherent Exciton States of Molecular Crystals, LBL-856, May 1972. -64- Alexander Pines, Principal Investigator various temperatures. Examples are shown in Figs. 1 and 2. Chemical shielding anistro- 1. NUCLEAR MAGNETIC DOUBLE-RESONANCE pies, which are difficult to obtain by any other means, have been reported for most of Alexander Pines these cases. When available, previous theories of chemical shielding, such as those of Two perennial problems in the study of Karplus and co-workers, fail to agree with dilute nuclear spins in solids by nuclear experimental results. Recent machine calcula magnetic resonance have been resolution and tions of Ditchfield are more successful for sensitivity. The observed spins may be dilute 13c. due to natural isotopic abundance OT due to chemical dilution. The observation of these The measurements provide much more com nuclei would open the way to many studies plete information than is available from not previously accessible to conventional liquid crystal experiments. In particular, nuclear magnetic resonance, these include, by using single crystals, complete informa for example, structural and dynamical fea tion on 13c shielding tensors has been ob tures of impurities in solids, microscopic tained in a number of compounds including details of nolecular motion in solids, phase durene, glycine, and oxalates. The direction transitions, spin relaxation phenomena, and of the principal axis of the shielding tensor quantan mechanical tunneling at low tempera in the molecule and the magnitudes of the tures. principal values are extremely sensitive to structure and bonding and should prove valu An approach to these problems has been made able aids in the study of electronic structure employing several well known features of solid in solids. state magnetic resonance. The problem of sensitivity is alleviated by transferring mag In addition to these structural studies, netic polarization from abundant surrounding several dynamical studies hav~ been made. •meleav spins (I) to the dilute nuclear spins These include high resolution examination of (S) under observation. In a typically useful I-S cross-relaxation and studies of molecular example the dilute spins could be S = 13c and the abundant spins I = lH. The polariza tion transfer can be accomplished in many ways, only one or two of which (consisting of transfer in the I, S double rotating frame) have been studied to date. In this way, the detected power signal/noise can be enhanced 2 over by a factor of G » (Y1/Y5) NT/Ng conventional Fourier transform spectroscopy. CY is the magnetogyric ratio and N the number of spins.) For *H - 13c G = 103 and for -H - -"N, another important case, G *= 10$. The problem of resolution is overcome by employing a solid state version of spin decoupling demonstrated by Sarles and Cotts, and by Mehring, Pines, Rhim, and Waugh. This involves intense irradiation of I spins by continuous or multiple-pulse rf fields. In this way, the resolution has been improved 100 200 for 13c from ~ 5 kHz to - 15 Hz. The tech Fig. 1. Proton-enhanced NMR spectrum of nique combining the sensitivity and resolution natural abundance 13c in polycrystalline enhancement has been termed "proton-enhanced acetic acid. The low field resonance is nuclear induction spectroscopy." from the carbonyl carbon and the high field one from the methyl group. The spectra The sensitivity gain has been verified in display characteristic shapes for anisotropy a .'.lumber of preliminary cases. For example, of the chemical shielding tensors. Principal 3 a high resolution spectrum of * C in - 50 mg values of the tensors can be read directly of solid adamantane was obtained in - 0.8 sec from the sharp edges of the powder peaks. with good signal/noise. Spectra have been The horizontal axis is in ppm relative to an 5 obtained in this way for 13c, 1 N, and 29Si external reference of benzene. in a large number of solid materials at (XBL 731-5627) -65- and this remains one of the most exciting aspects of these new techniques. Spectral changes have confirmed that in benzene and hexamethylbenzene, motion is facile about the molecular hexad axis as is known from wideline studies and neutron diffraction. Activation energies have been measured and are in agreement with accepted values. In adamantane, and in camphor, two "roundish" molecules, we have found that rotational motion is more or less isotropic in the solid, as may be expected. Rotational correlation times are ~ 10 9 sec at room temperature, suggesting rather serious intermolecular forces in the crystals. 2. NUCLEAR SPIN THERMODYNAMICS AND MULTIPLE- PULSE NUCLEAR MAGNETIC RESONANCE tff/, Wty*^Alexande r Pines In rigid solids, where nuclear spins are strongly coupled together by dipolar Fig. 2. 13C NMR spectrum in a single crystal interactions and only weakly coupled to the of durene obtained by proton-enhanced nuclear lattice, the spin system may be regarded for induction spectroscopy. Ten lines are many purposes as isolated and the concepts observed in this orientation of the crystal of thermodynamics and statistical mechanics reflecting the ten magnetically inequivalent applied. The spin temperature hypothesis carbons of the two molecules in the unit cell. states that after a spin-spin relaxation time In the liquid, only three lines are observed. T2, the system reaches internal equilibrium The most shielded component of the aromatic and is described by a density matrix shielding tensor is perpendicular to the Peq = (l/z) exp(-gH) appropriate to a ring. The horizontal scale is in ppm canonical ensemble. Many important phenomena relative to an external reference of benzene. have been understood and many elegant (XBL 731-2021) experiments have been devised on the basis of this theory. Two aspects of this have been looked at by us and studies will continue. motion. The complete dependence of cross- relaxation on rotating field strengths and In the first place, we may ask what happens frequencies is important for comparison with •when the Hamiltonian H becomes time dependent density matrix perturbation theories of due, say, to phase and amplitude modulated relaxation. Tne measurements yield details rf irradiation? In this case nuclear spin on fluctuation spectra of the cross-coupling thermodynamics no longer applies in the between nuclear spins. In the 19F - 43ca laboratory frame. We have shown, however, double-resonance in CaF2 studied by Hahn and that in a well-chosen "toggling" frame of co-workers, the spectrum was found to be an reference the Hamiltonian may be effectively exponential function of the rotating frame tine independent and thermodynamics applied. Zeeman splitting, implying a Lorentzian The thermodynamical quantities of interest correlation function for the cross-coupling such as heat capacity and entropy are obtained fluctuation. This is different from the from the density matrix employing the calcu Gaussian form normally assumed for these lated effective or average Hamiltonian. The functions. We have carried out preliminary latter is calculated using coherent averaging experiments which suggest that this behavior theory. This approach is a direct extension is quite general, but a substantial amount of the Redfield ideas of spin temperature in of work remains ;o be done in this area both the rotating frame. This phenomenon is on theory and on experiments. extremely valuable since it allows a modifica tion of thermodynamical parameters by care Molecular motion can be studied by fully monitored external perturbations. This observing both high resolution spectral should find applications to double-resonance changes and spin-relaxation times (cross- and to polarization enhancement in production relaxation and spin-lattice relaxation). of nuclear ferro- and antiferromagnetism. Again, only preliminary work has been done Preliminary experiments have been performed -66- on the 19F spins in a crystal of CaF^. Under was shown that whatever the irradiation, the a train of intense phase-alternated rf pulses average Hamiltonian for dipolar interactions applied at resonance the system was found to is just a (Van Vleck) truncated dipolar reach a state of internal equilibrium, and Hamiitonian along some axis in the rotating the subsequent increase in spin temperature frame. Recipes for calculating the relevant as a function of nutation angle agreed well parameters were given. It was shown that with theory. this should also be extremely useful for applications to line-narrowing, time-reversal, Secondly, it has been shown that the spin- and nuclear double-resonance in solids. temperature hypothesis must be employed with caution. If the sign of the Hamiltonian 3. RESEARCH PLANS FOR CALENDAR YEAR 1973 describing interactions between spins could be reversed, then the random phase approxima Alexander Pines tion which forms the basis for the selection of a canonical density matrix is no longer a. High Resolution Nuclear Magnetic applicable, and coherent phenomena could be Resonance in SolidT observed which would appear to violate the expectations of spin thermodynamics. Such Work will continue on techniques to an effective time-reversal has indeed been enhance both resolution and sensitivity. We 19 produced by Rhim, Pines, and Waugh in F plan specifically to develop ultra-high and lH spins in several solids, and spin sensitivity by employing indirect detection echoes following a free induction decay have of S spins by monitoring destruction of I been observed. The time-reversal is effected polarization in cross-relaxation. The by carefully selected multiple-pulse rf dynamics of polarization transfer will be 1 irradiation. An example on the ^F spins of studied and alternative transfer methods CaF2 is shown in Fig. 1. The free induction devised. We plan to investigate multiple decay in a solid had previously been thought adiabatic cross-over as a method for pro by some workers to be an irreversible process. ducing efficient total transfer of I to S The phenomenon has important application to polarization. The experiments will involve the study of line shapes, high resolution a large amount of technical work, and a nuclear magnetic resonance in solids, and versatile low temperature spectrometer is spin diffusion. being constructed. This wi 11 operate with an on-line minicomputer and superconducting The theory of the response of coupled solenoid, and will have facilities for nuclear spins to such irradiation must be well operation with multiple frequencies, high understood in order that these phenomena be power, and flexible timing in phase and exploited efficiently. The theory was ex amplitude modulation. The techniques will tended by Pines and Waugh to cover the general be used to extract information on structure case of irradiation far from resonance. It and motion in organic and "fluxional" organometallic compounds. In addition techniques will be developed to enhance sensitivity by strengthening the spin lattice coupling. This will be done by chemical doping and by optical pumping. Limitations posed by spin diffusion will be examined. b. Ferroelectric Phase Transitions Studies will be undertaken to obtain microscopic details on phase transitions in ferroelectric crystals of the KDP (potassium dihydrogen phosphate) class. We hope to Fig. 1. Time reversal of observe details of the transition in spectral 19F changes of the high resolution spectrum of free induction lH, Sip, 13c, etc. This should supplement decay in CaF2. Following a 90° pulse we information obtained from nuclear quadrupole observe the decay, following which a strong rf resonance studies of ^H and "As and help to perturbation is applied at the *9F resonance clfrify the role of hydrogen-bonding on the frequency. This effectively reverses the ferroelectric behavior in these cornpounds. sign of the nuclear spin dipolar coupling We plan to extend these studies later to and produces a spin echo when the spins are look at hydrogen bonding in solids exhibiting again allowed to propagate freely. The echo optical activity. maximum is ~ 500u sec from the initial 90° pulse. (XBB 731-278) -67- c. Optical-Nuclear Double-Resonance *5. M. G. Gibby, A. Pines, and J. S. Waugh, Proton-Enhanced Nuclear Induction Spectro Excited triplet states of organic molecules scopy. "Si Chemical Shielding Anisotropy in solids have been studied by Harris and in Some Organo-Silicon Compounds, J. Am. co-workers by phosphorescence-microwave Chem. Soc. 94, 6231 (1972). double-resonance. Electron spin echoes have been produced and are thought to decay due *6. A. Pines, M. G. Gibby, and J. S. Waugh, to modulation of nuclear-electron hyperfine Proton-Enhanced Nuclear Induction Spectro interactions by nuclear spin diffusion. Two scopy, l^c Chemical Shielding Anisotropy in types of nuclear magnetic resonance experi Some Organic Solids, Chem. Phys. Lett. 15, ments will be undertaken: 373 (1972). (i) To observe the nuclei near the excited *7. M. G. Gibby, R. G. Griffin, A. Pines and triplet molecule. This should yield J. S. Waugh, High Resolution NMR of 1SN in information on the local structure and Solids, Chem. Phys. Lett. 17, 80 (1972). information on resonance frequencies which is important for understanding *8. A. Pines and J. S. Waugh, Quantitative the electron spin relaxation. Aspects of Coherent Averaging. Simple (ii) To develop methods to inhibit spin Treatment of Resonance Offset Processes in diffusion effectively. This should Multiple-Pulse NMR, J. Mag. Res. 8_, 354 (1972). be useful later in extracting homo geneous electron spin line widths. *9. J. S. Waugh, A. Pines, and W-K. Rhim, Spin Echoes and Loschmidt's Paradox, J. Pure 4. 1972 PUBLICATIONS AND REPORTS andAppl. Chem. 32_, 317 (1972). Alexander Pines and Associates LBL report Journals 1. W. G. Breiland, C. B. Harris and A. Pines, Optically Detected Electron Spin *1. A. Pines, W.-K. Khim, and J. S. Waugh, Echoes and Free Precession in Molecular Homogeneous and Inhomogeneous Nuclear Spin Excited States, LBL-1412, December 1972 Echoes in Solids, J. Mag. Res. 6_, 457 0972). (submitted to Phys. Rev. Letters). *2. M. G. Gibby, A. Pines, W.-K. Rhim, and Papers presented J. S. Waugh, 31P Chemical Shielding Anisotropy in Solids. Single Crystal and Powder Studies *1. Proton-Enhanced Nuclear Induction Spectro at 99.4 MHz, J. Chem. Phys. 56, 991 (1972). scopy. Chemical Shielding and Nuclear Spin Relaxation of *3C in Solids, Thirteenth *3. A. Pines, M. G. Gibby, and J. S. Waugh, Experimental NMR Conference, Asilomar, Proton-Enhanced Nuclear Induction Spectro California, 1972. scopy. A Method for High Resolution NMR of Dilute Spins in Solids, J. Chem. Phys. 56, *2. Proton-Enhanced NMF. of Dilute Spins in 1776 (1972). Solids, Colloque Ampere, Turku, Finland, 1972. *4. M. G. Gibby, A. Pines, and J. S. Waugh, Anisotropic Nuclear Spin Relaxation of *-~C *3. High Resolution Nuclear Magnetic in Solid Benzene, Chem. Phys. Lett. 16, 296 Resonance in Organic Solids, Meeting of the (1972). NMR Discussion Group of the Chemical Society (London), Surrey University, England, 1972. Not supported by fflRD (work done before joining IMRD). -68- Ronald R. Uertn, Principal Investigator formed with broader CM cross sections and higher average product recoil energies, E', 1. COLLISION DYNAMICS OF "ALKALI-LIKE" ATOMS than are MCI; average E' values are ~ 8-12 kcal/mol for MI + CI and ~ 3-5 kcal/mol Boyd L. Earl, Lara A. Gundel, Ronald R. Herm, for MCI + I. Shen-Maw Lin, Lambert C.-H. Loh, Charles A. Mims, and Charlotte M. Sholeen Ba, Sr, and Ca + Some Halides of Methane-- This same apparatus Has also been used to Our past research efforts have concentrated measure angular distributions of the following on experimental studies of the collision mass filter detector signals: MI+ from dynamics of gaseous atomic species with Ba, Sr, and Ca + CH3I and CH2I2; Bal+ from relatively low ionization potentials. CF3I; and BaCl+ from Ba + CCI4. The results Research during 1972 which was directed along show that the HI + CH3 products from M + CH3I these lines included: (1) measurements of scatter predominantly backwards in the CM the cross sections for collisional quenching system with substantial recoil energy (average of excited alkali atoms produced by photo- E1 value * 10-15 kcal/mol); these results dissociation of alkali halide vapor (BLE); are very similar to those observed for the (2) a study of the electronic chemilumines- analogous alkali atom reactions. All of the cence from HgCl which is produced in a flowing other reactions preferentially scatter prod Hg-Cl? vapor mixture upon irradiation by the ucts into the forward CM hemisphere. For 2537 A Hg resonance line (LAG); (3} crossed these reactions, the MX* mass filter signal beam studies of the reactions of Li atoms with might arise from ionization of a mono- or di- halogen-containing molecules (CMS and LAG) ; halide product. However, both the CH2I2 and (4) crossed beam studies of the collisional CF3I reactions show striking differences from quenching of Hg (6s6p3Pn 2) (LC-HL); and those of the analogous alkali reactions: for (5) crossed beam studies'of some reactions Ba • CF3I, a weak BaF* mass filter signal is of Ba, Sr, Ca, and Mg atoms (S-ML and CAM). also observed; and the MI* LAB angular dis Since our 1972 efforts were an extension of tributions from Ba, Sr, and Ca+ CH2I2 are earlier studies described in previous IMRD only slightly broader than the calculated Annual Reports, most of this work will not angular distributions for the center-of-mass be described in this present report. However, velocity vector. These differences are quite it seems worthwhile to include the following reasonable if one assumes that MI2 (and description of our results for some of the possibly MI as well) forms in the CH.jIj alkaline-earth atom reactions because this reactions and that BalF (and probabl' Bal as work is mentioned only briefly in the 1971 well) is formed from Ba + CF3I. Annual Report. Ba, Sr. Ca. and Mg + IC1 and Ba BrCN*— * Abstracted from LBL-1134. Primitive product angular distributions of MCI and MI from Ba, Sr, Ca, and Mg + IC1 and of BaCN and BaBr from Ba + BrCN have been measured in a crossed-beam apparatus employ 2. CROSSED BEAM REACTIONS OF BaF WITH BCI3 ing an electron bombardment ionizer-mass fil 2 ter detector. Product center-of-mass (CM) Richard Behrens, Jr., Andrew Freedman, recoil angle and energy distributions have Timothy P. Parr, and Ronald R. Herm been fit to the measured laboratory (LAB) angular distributions by averaging the CM •> LAB transformations over the (non-thermal) The development of the electron bombard beam speed distributions. Both the MCI and ment ionizer-mass filter "universal" detector MI CM product distributions from M + IC1 makes possible crossed beam studies of the are relatively insensitive to the identity reaction dynamics of a variety of high tem of the attacking atom, except that practical perature species. Thus, we have recently ly no Mgl is formed in the Mg reaction. Other examined the reactive scattering from crossed beams of BaF (at 1S50°K) and BCl (at 340°K). product yield ratios are estimated as — 2:1 2 3 We have observed BFC1+ and BF + mass filter for MC1:MI from 'IC1, ~ 10:1 for BaCN:BaBr 2 signals of reactively scattered species. from BrCN. All of the reactions favor for + ward CM product scattering (i.e., MX scattered (A BaCl reactive scattering signal cannot be distinguished from a BaF2+ non-reactive within 90° of the direction defined by the + incident M), although the preference is signal.) The BF2 signal must be due to the especially weak in the Ba + BrCN * BaBr + CN reaction: reaction. For the IC1 reactions, MI are BaF2 + BCI3 ->• KaCl2 + BF2C1 . (1) -69- The BFC1* signal might arise from BF2C1 from molecule-molecule reactions, such as reaction (1] or from BFCI2 from: the BaF2 + BCI3 reaction described in the previous section, will also be sought. How BaF2 + BCI3 - BaFCl + BFCI2 . (2) ever, most of our efforts will be directed towards the study of reactions of molecular + free radicals; initially, the reactions of We have not observed signals due to BaFCl , + + + the OH radical will be studied. These studies BaCl2 , BF2C1 , or BFCl2 ; however, this is are likely to be difficult because of problems not surprising, as the product molecules in the generation of an intense radical beam. would be expected to fragment upon electron However, for certain radicals, such as OH, bombardment ionization. it should be possible to overcome this dif ficulty by operating the beam source at suf Results are still preliminary and center- ficiently high temperatures so that the free of-mass product distributions have not been radical becomes a major, or dominant, equili obtained because of an undetermined "viewing brium species. Another theme which will be factor" distortion in the measured laboratory explored envisions the use of the phenomenon angular distribution. However, we would of condensation in expansion from a high emphasize that our beav. intensities are suf pressure source as a method of preparation ficiently low to ensure "single collision" of unusual molecules for collision dynamics conditions. Furthermore, we estimate that studies. Possibilities here include the our apparatus sensitivity is such that the study of reactions of weakly bound "van der observed BF2 signal implies a total reaction Waals" molecules (e.g., A^). Our progress cross section, QR, of at least 1 A2 for in these areas during 1973 will be dependent reaction (1). This study was originally on available funds, so that we may seek undertaken because it was anticipated that outside support as well. QR > 1 A2 for reaction (25 because of the likely stability of the intermediate collision complex. In analogy with the alkali halide- boron trihalide adducts, this complex was expected to have an (BaF)+ (BCI3F)" structure 4. 1972 PUBLICATIONS AND REPORTS and to dissociate into the products of reac tion (2). The occurrence of reaction CI) Ronald R. Herm and Associates was surprising and might be due to a "secondary complex" formed as BaFCl and BCI2F Journals attempt to separate; the dipole-dipole attraction between these products might favor 1. Boyd L. Earl, Ronald R. Herm, Shen-Maw such snarled secondary encounter trajectories. Lin, and Charles A. Mims, Photodissociation of Nal Vapor and the Energy Dependence of the Quenching of Na*(3p2p) by Foreign Gases, J. Chem. Phys. |6_, 867 (1972) (LBL-102). 3. RESEARCH PLANS FOR CALENDAR YEAR 1973 2. Charles A. Miirs, Shen-Maw Lin, and Ronald R. Herm, Crossed Beam Collision Ronald R. Herm Mechanics: Reactions of Ca, Sr, and Ba with HI and Limits on D0° for Cal, Sri, and Bal, Crossed beam studies of the reaction J. Chem. Phys. 57, 3099 (1972) (LBL-489). dynamics of Li r id Na atoms will continue through 1973 ir^o 1974. Our other studies LBL reports grouped under che heading of "alkali-like" atoms in this report should be completed 1. Shen-Maw Lin, Molecular Beam Kinetics during 1973. During 1973, we also expect to complete the final assembly of our more (Ph.D. thesis), LBL-891, September, 1972. sophisticated "universal detector" crossed beam apparatus, which was described ir. the 2. Shen-Maw Lin, Charles A. Mims, and 1971 Report. Ronald R. Herm, Crossed Beams Chemistry: Reactions of Ba, Sr, Ca, and Mg with CI2 and Br2, LBL-1126, August, 1972 (submitted to We will seek out new themes for our crossed beam studies of collision dynamics during J. Chem. Phys.). 1973. One of these will be studies of the 3. Charles A. Mims, Shen-Maw Lin, and reaction dynamics of high temperature species. Ronald R. Herm, Crossed Beam Product Angular This will continue to include some studies Distributions: MCI and Ml from Ba, Sr, Ca, of reactions of metal atoms. Thus, a few and Mg + IC1 and BaCN and BaBr from reactions of alkaline-earth atoms will be Ba + BrCN, LBL-1134, October, 1972 (submitted examined in finer detail; reactions of to J. Chem. Phys.). Group III A atoms (notably Al) will also be studied. Further examples of bimolecular -70- 5. Shen-Maw Lin, Charles A. Mims, and Ronald R. Herm, Crossed Beams Reactions of Ba, Sr, and Ca with Some Halides of Methane, LBL-1175, December 1972. (submitted to J. Phys. Chem.) -71- Harold S. Johnston, Principal Investigator 1 1 ' I ' 1% Introduction. In previous years we have • io-" - developed a new method for observing the spectra and chemical kinetics of reactive | [ intermediates [for example, free radicals) produced photochemically by relatively low intensity light sources. Within the last two ! years it has become apparent that this work io-'» 1 is pertinent to use of the stratosphere by supersonic transports. In 1972 our regular V"— •• ! l .a> program was substantially expanded by a 0 s contract from the Department of Transporta .'/ s ; tion. Our contract is one of 50 in the / IO-" / country supported by the Climatic Impact / Assessment Program of the Department of / \ ; / \\ Transportation. The overall goal of the ,n-*= • \ national program is to assess the properties, 190 210 230 250 270 290 310 330 resources, and limitations of the stratosphere Nanometers as a region for transportation. Our project Fig. 1. Graph showing the room temperature concerns the spectra and photochemistry of the oxides of nitrogen, free radicals based spectrum of nitric acid vapor. XBL 726-6417) on water CH, HD, HOO), and ozone. Nitric acid is destroyed by hydroxyl radicals: 1. EXPERIMENTAL STUDIES a. Visible and Ultraviolet Absorption Spectra HO + HNO3 -* H20 + NO3 (2) of Oxides of Nitrogen and Oxy-Acids of Nitrogen and by photolysis by ultraviolet radiation: Richard Graham and Harold S. Johnston HNO3 • hv -• HO + N02 The goal is to obtain accurate, quantita tive absorption spectra at temperatures and * H + NO3 (?) . (3) pressures close to stratospheric conditions (1 to 100 Torr, 200-300°K) for the following The steady-state ratio of nitrogen dioxide to gases: N0 , NO3, N2O3, N2O4, N2O5; HNO3, 2 nitric acid is HNO2; O2, O3. The room temperature spectrum of nitric acid vapor is given as Fig. 1. The [N0 ] k room temperature spectra of NO2, NO3, N2O5, 2 \ 3 + and HNO2 have been obtained, but the work is TRSE33TS "kY k^TB0T • not yet complete. A major effort has been devoted to converting the long path (32 meter, ultraviolet; 50 meter, infrared) spectrometers The rate constants of reactions (1) and (2) for use at low temperatures. The converted are known, and the rate of photolysis of instruments are to be used for kinetic studies nitric acid is needed. and for photochemical reactions as well as for spectroscopic measurements. The photolysis of nitric acid vapor is a very complex laboratory reaction. Nitrogen b. Quantum Yields for the Photolysis of dioxide is photolyzed by the same radiation Nitric Acid Vapor" that photolyzes nitric acid vapor. The product, NO, undergoes surface-catalyzed Ted Chang and Harold S. Johnston reaction with nitric acid. We are studying this photochemical reaction with a variety In the stratosphere, nitric acid is formed of experimental methods. The goal is to from hydroxyl radicals and nitrogen dioxide: identify the primary photolysis products and the quantum yield as a function of wavelength. Results obtained at 210 nm and at 290 nm HO + NO2 HNO3 (1) indicate a primary quantum yield of about 1, -72- and the primary product is probably HO (it presence of oxygen, the "odd chlorine" is in must be either H or HO to account for observed the forms: CI, CIO, C100 (peroxy form). We products). have studied the kinetics of the ClOO and ClO radicals, the ultraviolet spectra of CIO c. The Rate of Reaction of Nitrogen Dioxide and ClOO, and the infrared spectra of C100. with Ozone The addition of carbon monoxide to this system leads to additional ClO: T. T. Paukert, John Birks, and Harold S. Johnston ClOO + CO •* CIO + C02 . The fastest catalytic cycle for ozone destruction by the oxides of nitrogen under In an effort to obtain the infrared spectrum stratospheric conditions is the NO2 cycle: of ClO, we added carbon monoxide. The infrared spectrum of CIO was probably detected, but + it is partly obscured by a strong radical NO + O3 -<• NO2 O2 spectrum, perhaps C100CO or 00(00)CI. This 1*>2 + 0 •* NO f O2 study has been completed and is being written up for publication. net: 0 * O3 •*• O2 + O2 f. The Quantum Yield of Singlet Oxygen from Oxygen Photolysis' There is another cycle that may be very important below 22 km: Douglas R. Martin and Harold S. Johnston NO2 + O3 ->• NO3 + O2 When ozone absorbs radiation throughout the visible and ultraviolet region, the NO3 + hv •> NO + O2 (day) primary photochemical reaction is: NO + O3 •>• NO2 + 02 03 + hv •* 02 • 0 . net: 2O3 + red light •* 302 However, the electronic state of the oxygen atom product varies with wavelength: below The key reaction in the NO3 cycle is that 290 ran it is O^D), above 320 nm it is 0(3P), between NO2 and ozone. This reaction has but between 290 and 320 nm the quantum yield been observed only at room temperature and for 0C D) drops from one to zero. The rate over a very narrow temperature range. The of formation of OC^D) in the troposphere and temperature range is entirely too narrow to lower stratosphere is very sensitive to how support an extrapolation to stratospheric the quantum yield for formation of 0(1D) temperatures. This reaction is being studied varies with wavelength. in a cell with 3 meter optical path and with temperature controls down to 230°K. Results We have under way a study of the yield are being obtained, but the study is not yet complete. of 0(4>) as a function of wavelength for narrow (1 nm) bands between 290 and 330 nm. d. The Photolysis of Nitrogen Dioxide g. Fluorescence of Sulfur Dioxide Alan B. Harker and Harold S. Johnston Douglas R. Martin and Harold S. Johnston Although the photolysis of nitrogen dioxide has been studied by many different The dynamics of vibrational energy trans investigators, there has remained some fer in highly excite." molecules can, in disagreement on important aspects of the favorable cases, be studied by means of study. Also the transient appearance and fluorescence studies. We have studied the decay of nitrogen pentoxide is of some fluorescence of SO2 in a large (22 liter) interest. A study of this reaction has been spherical Pyrex bulb equipped with silica completed and an article submitted for windows. The wavelength distribution of publication. Our results are in good agree fluorescence radiation as a function of ment with one group of literature values, but pressure (around 1 micron) has been followed disagree with the results of one investigator. for various narrow bands of exciting radiation. This study has been completed, and the e. The Free Radicals CIOY preparation of a thesis and journal article is in progress. Alan B. Harker and Harold S. Johnston When chlorine atoms are produced in. the -73- h. Reactivity of Hydroxyl Radicals as a total ozone since 1960. For a small number Function of Vibrational Energy State of these stations (10 out of about 90), Komhyr, Barrett, Slocum, and Weickmann Jere E. Streit and Harold S. Johnston [Nature 232, 390 (1971)] found that over the 10 year period 1961-70 (inclusive) ozone Hydroxyl radicals, HO, are produced in a appeared to increase by 2 to 104. In the large, low-pressure apparatus from the reaction spring of 1972 we started a statistical study of hydrogen atoms and nitrogen dioxide: of all the data, using the procedure of Kbmhyr et al., with some improvements in certain H + N02 * HO + NO . details. We expected to find as many stations showing decreases as Kbmhyr found with increases, and we expected the global average The hydroxyl radicals are produced in excited change to be zero within experimental error. vibrational states, which are observed by To our surprise we confirmed the previous means of emissions spectra. The rate of study and found an average increase of ozone reaction of vibrationally excited HO radicals of 4.5*1.2 (20) percent per decade (1960-70) with ozone is being followed. for all stations. The increase was larger for the period 1963-70. For the period 1963-70 the increase was 4.6*1.4$ for all stations; it was 5.6*1.5% in the northern hemisphere 2. THEORETICAL STUDIES and 7.6*3.2% for stations north of 50°N. On the other hand, ozone decreased during the a. Instantaneous Global Rates of Formation three years 1960-62 at a percentage rate per and Destruction of Ozone in the Stratosphere decade of: -7.1*6.5 (2o) for all stations, -7.6*7.0 for northern hemisphere, and Gary Z. Whitten and Harold S. Johnston -12.6*10.6 above S0°N. For more than 10 years atmospheric scientists have recognized that ozone is These changes as a function of space and produced from sunlight much faster than it time are consistent with a hypothesis that the is destroyed by reactions involving pure decreases (1960-62) were due to nitric oxide oxygen and nitrogen. "Something else" is injected into the stratosphere by nuclear very active in destroying ozone in the natural bomb tests (1952-62), and that the increase stratosphere. The role of air motions in the of ozone (1963-70) was the world slowly stratosphere make very difficult the evaluation returning to normal after the cessation of of the magnitude of the role of "something nuclear tests in 1962. else." We have devised a relatively simple computational procedure that gives the fol c. ActivationEnergies for the Dissociation lowing balance sheet for the global natural of Diatomic Molecules stratosphere for rates per second Cany second) in units of 1029 molecules of ozone: John Birks and Harold S. Johnston Ozone formation +S00 An analysis of all data on the dissociation of diatomic molecules showed that the observed Loss in "pure air" -86 decrease of activation energy (to values well Transport to troposphere -6 below the dissociation energy) with increasing temperature can be explained in terms of non- "Something else" -408 equilibrium distributions over vibration levels of the molecule and in terms of allowed Thus the mysterious "something else" accounts direct decomposition from any and all for over 80? of the natural removal of ozone. vibrational states. We find 4*ip9 molecules NO* (NO and N02) per cm3 in the stratosphere to be sufficient to account for the needed 408 units of ozone loss. However, some of this loss is probably 5. RESEARCH PLANS FOR CALENDAR YEAR 1973 brought about by water vapor, especially in the upper half of the stratosphere. Harrld S. Johnston b. Statistical Studies of "Ozone Data for During calendar 1973, we plan to study the World. 1960-70" the absorption spectra of several oxides of oxyacids of nitrogen at temperatures comparable Gary Z. Whitten and Harold S. Johnston to the stratosphere, to study quantum yields (HNOSJ 03I N03) as a function of wavelength, A large network of stations has regularly to study the rate of reaction of NO2 with reported daily and monthly measurements of ozone, and to study several photochemical -74- reactions involving free radicals based on 7. H. S. Johnston, in The Oxides of Nitrogen water (H, HO, HOO). The theoretical studies and Photochemical Smog, Atomospheiric will focus on further identification of the Chemistry and Air Pollution (Air Resources important reactions in the stratosphere. As Center, Oregon State University, 1972) p. 69. others determine the actual distribution of NOx in the stratosphere, we are prepared to Papers presented calculate whether or not this amount of N02 is sufficient to be the "something else" that 1. H. S. Johnston and Gary Whitten, now removes 80% of the naturally produced Instantaneous Photochemical Rates in the ozone. Global Stratosphere, presented at the International Ozone Symposium, Arosa, Switzerland, August 21-25, 1972 (LBL-1178). 4. 1972 PUBLICATIONS AND REPORTS 2. H. S. Johnston and Gary Khitten, Sensitivity Tests for Sufficiency of the Harold S. Johnston and Associates Chapman Reactions, presented at the Meeting of American Geophysical Union, San Francisco, Journals and book December 5, 1972. 1. K. S. Johnston, The Effect of Supersonic 3. H. S. Johnston, Laboratory Chemical Transport Planes on the Stratospheric Ozone Kinetics as an Atmospheric Science, Presented Shield, Environmental Affairs !L, 73S (1972). at Survey Conference on Climatic Impact Assessment Program, U.S. Department of 2. T. T. Paukert and H. S. Johnston, Spectra Transportation, Cambridge, Mass., Feb. 15-16, and Kinetics of the Hydioperoxyl Free Radical 1972. (LBL-497.) in the Gas Phase, J. Chem. Phys. 56, 2824 (1972)(UCRL-19109 Rev.). — LBL reports 3. H. S. Johnston, The Concorde, Oxides of 1. A. B. Harker, Reaction Kinetics of the Nitrogen, and Stratospheric Ozone, Search 4, Photolysis of NO2, and the Spectra and 276 (1972). Reaction Kinetics of the Free Radicals in the Photolysis of the CI2-O2-CO System (Ph.D. 4. H. S. Johnston, Newly Recognized Vital thesis), LBL-1114, September 1972. Nitrogen Cycle (Oxides rf Nitrogen/ Stratospheric Photochemistry/Ozone Balance), 2. Harold Johnston and Richard Graham, Gas Proc. Nat. Acad. Sci. 69, 2369 (1972). Phase Ultraviolet Absorption Spectrum of Nitric Acid Vapor, LBL-1179, October 1972. 5. C.-H. Wu and H. S. Johnston, Molecular Modulation Mass Spectrometry Kinetic Study 3. Harold Johnston and Gary Whitten, of the CIO Free Radical, Bull. Soc. Chim. Instantaneous Photochemical Rates in the Beiges 81, 135 (1972). Global Stratosphere, LBL-1178, October 1972. 6. J. Birks and H. S. Johnston, Activation Energies for the Dissociation of Diatomic Molecules Calculated from Molecular Properties, Ace. Chem. Res. £, 327 (1972)(UCRL-20514 Rev.). -75- Bruae S. Mahan, Principal Investigator 0?*!%— PjO'. Bruce H. Mahan The mutual neutralization of gaseous ions is one of the important mechanises by which an ionized gas can decay to neutrals. The process can occur by parallel bimolecular (two ion] or termolecular (two ions, one neutral) reactions, and the relative impor tance of these mechanisms depends on the total gas pressure. At pressures above one atmosphere and at room temperature, the rate of recombination by either mechanism is limited by the rate at which ions diffuse toward each other. We have reviewed the experimental and theoretical work on gaseous ion recombination, much of which was done in our laboratory. The general reliability of the data has been Fig. 1. Contour maps of the specific intensi assessed, and comparisons made with various ty of O2D* from the O^-Cz^z reaction at theories. The rate coefficients for both 125 and 99.4 eV laboratory energy. The small bimolecular and termolecular processes can crosses locate the velocity of O2D* formed in fact be calculated from the properties of by the spectator stripping process. The the reactants and possible products to circles marked Q = -2.3 eV and Q = -10 eV within an order of magnitude, which is a are, respectively, the velocity limits for uniquely satisfactory situation in chemical products with the minimum and maximum allow kinetics. Nevertheless, discrepancies be able internal excitation consistent with tween theory and experiment suggest several their stability in the ground electronic ideas and experiments which should be pursued. state. The angles are measured in the barycentric system. (XBL 721-5967) Abstracted from a manuscript to be published in Advances in Chemical Physics, 1973. fore been associated with exoergic processes exclusively, and the experiments reported here demonstrate that spectator stripping + 2. THE 02 -C2D2 PACTION: AN ENDOERGIC can occur for endoergic reactions as well. STRIPPING PROCESS An energy threshold for the abstraction MingHwa Chiang, Bruce H. Mahan, and occurred at a laboratory energy of approxi Charles Maltz mately 40 eV for the 02+ projectile/ At this laboratory energy, O2* has an energy relative We have determined the product D02+ veloci to the entire C2D2 molecule of 17.7 eV, much ty vector distributions for the endoergic in excess of the endoergicity of 2.23 eV. atom abstraction reaction However, the energy of Oz+ relative to the atom it abstracts is close to 2.23 eV at the °2+ + C2D2 •* C2D + DC2+ AH - 2.23 eV reaction threshold. This is a further indica tion of the spectator nature of the C2E frag ment in the reaction. at a number of values of the initial relative energy. Two examples of these distributions are shown in Fig. 1. From these product velocity vector maps, it is clear that the Abstracted from a paper to be published in abstraction reaction takes place by a specta J. Chem. Phys. and LBL-1101. tor stripping mechanism in which little or no momentum is transferred to the C2D fragment. This type of mechanism had hereto -76- 3. DYNAMICS OF THE 0+-H? REACTION.I: REACTIVE SCATTERING OF 0*(4S3/2) AT RELATIVE O'+HD — OH* + D l75eV! t, ENERGIES BELOW 15 eV Relative Energy = ll.9eV | Keith T. Gillen, Bruce H. Mahan, and John S. Winn We have made extensive measurements of the ion product velocity vector distributions for the reaction 0+(H2,H)0H+ and its isotopic variations. Figure 1 shows a contour map of the product intensity in the barycentric system which clearly indicates that the reaction proceeds by a direct interaction 20% mechanism, rather than through a long-lived Beam H20+ intermediate complex. Spectator strip Profile ping is a prominent feature of this and other product distributions obtained at low (< 8 eV) relative energies of collision. Figures 2 and 3 show, respectively, the OH* and 0D+ obtained from 0+-HD collisions at a relative energy of 11.9 eV, and illustrate a number of the features which characterize the reaction at high relative energy. In Fie. 2. A contour map of the intensity of both instances, the peak at the spectator OH* formed from 0+-HD collisions at 11.9 eV stripping velocity has disappeared, since at relative energy. Note the absence of an these energ^js ground state 0H+ and 0D+ intensity peak at the stripping velocity formed by spectator stripping would be suf (marked by a small cross), the substantial ficiently internally excited to be unstable intensity at velocities inside the Q = -4.5 eV. with respect to dissociation. The propensity circle, and the high intensity at angles less for OH* to be formed at angles less than 90°, than 90°. (XBL 727-6588) and OD* to occur at angles gioiter than 90° is a consequence of general kinematic con- 0*+HD—»OD +H Relative Energy = H.9eV 0* + Hj — OH** H MOeV) Relative Energy = 1.5 eV Fig. 3. A contour map of the intensity of Fig. 1. A contour map of the intensity of 0D+ formed from 0+-HD collisions at a rela Or from the O^-H? reaction at 4.5 eV rela tive energy of 11.9 eV. Note the propensity tive . leigy. The small cross locates the for backscattering, that is, formation of velocity of OH* formed by the spectator products at angles of 90° or greater. strippir^ process. (XBL 727-6542) (XBL 727-6593) -77- siderations which apply to collinear or from free D atoms. Similar results are found nearly collinear collisions. The appearance for the other isotopic targets H2 and HD. In of some product in regions inside the the latter case, two inelastic features can Q = -4.5 eV circle, where ground state OH+ be discerned in the high energy experiments, or OD+ would be unstable with respect to one coming from scattering of apparently free dissociation, is indicative of the formation H, the other from apparently free D. of OH* fa) or OD+C1A) in some fraction of the collisions. These phenomena indicate that at energies above 15 eV, the collisions of 0+ with H2, D2, and HD are essentially impulsive in Abstracted from a paper submitted to J. Chem. nature. Therefore, both the non-reactive and Phys., and LBL-1429. reactive scattering should be described to a good approximation by a model in which the three particles interact via hard sphere potential functions. We are currently + exploring the predictions of this model, and 4. DYNAMICS OF THE 0 -H2 REACTION. II: the initial results show encouraging agree REACTIVE AND NON-REACTIVE SCATTERING AT ment with experiment. RELATIVE ENERGIES ABOVE IS eV Keith T. Gillen, Bruce H. Mahan, and John S. Kinn 5. AN APPARATUS FOR THE STUDY OF ION- We have studied the reactive and non- MOLECULE REACTIONS AT LOW RELATIVE COLLISION reactive scattering of 0+ from H2, D2, and ENERGIES HD at relative energies of collision ranging from 15 to 50 eV. A most striking feature James A. Fair and Bruce H. Mahan appears in the non-reactive scattering, as is illustrated in Fig. 1. While the small- We have constructed an apparatus which angle scattering (8 < 45°) has a clear allows us to determine the velocity vector elastic component, the scattering at larger distributions of the ionic products of ion- angles is very inelastic. In particular, molecule reactions carried out with ion the intensity ridge found at 70° and larger projectile energies between 5 and 20 eV in angles lies approximately on a circle which the laboratory system. This apparatus thus would be the locus of elastic scattering allows us to extend the range of our collision dynamics experiments to significantly lower energies than has been previously possible in our laboratory. The apparatus consists of a small 90° mag netic mass spectrometer to prepare the primary ion beam, a scattering cell or crossed supersonic molecular beam to provide the tar get molecules, and a Men filter-quadrupole mass filter combination for the velocity and mass analysis of the products. Our first experiments on the N+(H2,H)NH* reaction indicates that this process occurs largely by a direct interaction mechanism even at relative energies as low as 1 eV. 6. RESEARCH PLANS FOR CALENDAR YEAR 1973 Bruce H. Mahan Fig. 1. A contour map of the intensity of We will continue our studies of the reac 0* scattered from D2 at an initial relative + + + 1 energy of 20.1 eV. All scattering inside tive scattering of simple ions (0 ,N ,C ,C02' ') the circle labeled Q = -4.5 eV is inelastic by small molecules in an effort to learn how enough to cause dissociation of the hydrogen. to anticipate reaction dynamics from the The circle labeled elastic 0+(D,D)0+ gives the properties of reactants and products. Studies locus that 0+ would have if it had been of the reactions of metastable electronically scattered off free D atoms. (XB1 727-6587) excited ions have begun, and will be con tinued. Investigations of the reactions of -78- metallic ions will be initiated. We have 2. Bruce H. Mahan and John S. Winn, Inelastic begun the design and will continue the Scattering of Ne+ by H2 and D2, J. Chem. development of a fast cross-correlation Phys. 57_, 4321 (1972) (LBL-1117). method for the velocity analysis of ion- molecule reaction products. All these projects 3. Ming-Hua Chiang, Bruce H. Mahan, and are aimed at increasing our general under Charles Maltz, The 02+-C2D2 Reaction: An standing of chemical reactions, and specifical Endoergic Stripping Process, J. Chem. Phys. ly at clarifying the higher energy charged 57., 5114 (1972) (LBL-1101). particle collision phenomena which are impor tant in discharges, atmospheric and re-entry 4. Bruce H. Mahan, Recombination of Gaseous phenomena, and radiation chemical systems. Ions, Advan. Chem. Phys. 23, 1 (1973). Paper presented 7. 1972 PUBLICATIONS AND REPORTS Bruce H. Mahan, "Dynamics of Ion-Molecule Reactions," Gordon Research Conference on Brace H. Mahan and Associates Plasma Chemistry, September 1972. Journals LBL reports 1. Keith T. Gillen and Bruce H. Mahan, 1. Mervyn Ming-Bra Chiang, Reactive and Deconvolution of Molecular Beam Inelastic Inelastic Scattering of Molecular Ions, Scattering Data, J. Chem. Phys. 56, 2517 (Ph.D. thesis), University of California, (1972). Berkeley, LBL-474, February 1972. -79- William H. Miller, Principal Investigator where E(n) is the inverse function of n(E) from Eq. (1). 1. SEMICLASSICAL QUANTIZATION OF NON- SEPARABLE SYSTEMS Because of the difficulty of applying the above result, a semiclassical perturba William H. Miller and Jimmie D. Doll tion expansion for the eigenvalues of an N-dimensional system was carried out;2 The utility of the semiclassical, or WKB, explicit expressions were obtained for the eigenvalue relation for one-dimensional first- and second-order perturbation energies. eigenvalue problems is well known. Not only does it provide a useful approximation for the energy levels of a potential well (cf. 1. W. H. Miller, J. Chem. Phys. 56, 38 the Dunham expansion for vibrational — eigenvalues), but it can also be "inverted" (1972). (the Rydbsrg-KLein-Rees method) so that the 2. J. D. Doll and W. H. Miller, J. Chem. potential function itself can be determined Phys. S7, 4428 (1972). if the energy levels are obtained experimen tally (e.g., from spectroscopy). In light of this success of the semi- 2. PENNING IONIZATION OF HYDROGEN ATOMS BY classical approximation for systems with METASTABLE HELIUM ATOMS* only one degree of freedom, and also in view of its usefulness in scattering problems, W. H. Miller, Charles A. Slocomb, and it seems worthwhile to explore the possibility Henry F. Schaefer of extending it to systems with several non- separable degrees of freedom. The desire Detailed calculations have been carried is to facilitate the arilysis of the out for the Penning and associative ioniza eigenvalue spectrum of multidimensional tion of ground state H atoms by Is2s->S systems, the ultimate goal being the metastable He atoms, establishment of "inversions relations" whereby one could use the experimentally observed vibrational spectrum of a triatomic i | i " I - molecule, for example, to construct its multidimensional potential energy surface. -2.65 The first of these goals has been accom " plished, i.e., the generalization of the -2.70 WKB quantum condition to general non-separable 1 /i») systems. The semiclassical eigenvalue • V/ relation for a system of N degrees of freedom is related to the periodic classical -2.75 - trajectories of the system: if there are no constants of the motion other than the total eneTgy itself, then there is typically -2.80 just one periodic orbit for each energy E, - - and the quantum condition is -2.85 - n(E) + L/4 h'1 { dt E PjCt) qjCt). (1) Jo i-1 -2.90 / - In Eq. (1) {qj.(t), Pi(t)> are the coordinates /Vt(R) and momenta along the periodic trajectory of energy E, T is the period of this orbit, -2.95 h is Planck's constant, and L is a specifical ly determined integer. Equation (1) defines 1 , 1 1 : 1 the function n(E) for all values of the total 4 6 energy E in terms of the action integral R (Bohrs) along the periodic trajectory of energy E; Fig 1. The potential energy curves for the eigenvalues are ECn), n " 0, 1, 2, ..., He*-H(z£) and He-H+Clz). (XBL 716-6862) •*». 1 1 1 1 1 v. He* + H - He + If + e" CD 2 10"- - \ - -• HeH* + e' . (2) Figure^l shows the potential energy curves for He -H and He-tf that were calculated 1 4 previously, and Fig. 2 shows the autoioniza icr- v tion width r as a function of the He -H distance. (r has units of energy; the autoionization rate is r/h.) ^X From the quantities in Figs. 1 and 2 it -bD \ I0" - \ is possible to calculate the cross section \ for reactions CI) and (2) and the energy \ distribution of the ionized electrons as well.2 Figure 3 shows the total ionization V cross section [the sura of cross sections for V 10- 8 1 1 , 1 1 reactions CI) and C2)] and the associative 6 8 part separately, and Fig. 4 shows explicitly R (Bohrs) the associative fraction as a function of Fig. 2. The autoionization width of the the collision energy. metastable state He*-H as a function of intemuclear distance. QCBL 716-6860) Recent measurements of the associative ionization cross section by Neynaber* give an energy dependence of the cross section just at that in Fig. 3 and an absolute cross section of 0.22 A2 at 1 eV collision energy; as seen from Fig. 3, this is in essentially exact agreement with our calculation. Measurement of the energy distribution of the ionized electrons by Hotop et al.* also shows good agreement with our results for this quantity. Abstracted from J. Chem. Phys. 56, Ybr.l C1972). ~ O.OI 0.03 0.1 0.3 I E (ev) 1. W. H. Miller and H. F. Schaefer, J. Chem. Fig. 3. The solid curves are the total Phys. 53, 1421 (1970). ionization cross section Cupper curve) from 2. W.TT. Miller, J. Chem. Phys. 52, 3562 reactions (1) and (2), and the associative C1970). — ionization cross section (lower curve) 3. R. H. Neynaber, private communication. separately [reaction (2)], as a function of 4. H. Hotop, E. Illenberger, H. Morgner collision energy. (XBL 716-6859) and A. Niehaus, Chem. Phys. Lett. 10, 493 (1971). — 3. ANALYTIC CONTINUATION OF CLASSICAL MECHANICS FOR CLASSICALLY FORBIDDEN PROCESSES IN INELASTIC AND REACTIVE SCATTERING Thomas F. George, Jimmie D. Doll, and William H. Miller 0.1 A particularly important aspect of the E (eV) semiclassical theory of molecular collisions Fig. 4. The percentage of the total ioniza is the ability to extend or analytically tion crosi section that >iultls associated continue classical mechanics in such a way product HeH*, as a function of relative as to describe collision processes that lie collision energy. (XBL 716-6858) beyond the framework of ordinary classical -81- mechanics. The most familiar example of such a process is tunneling of a particle through a one-dimensional potential barrier, so that "classically forbidden" transitions in systems with more than one degree of freedom are a generalization of one- dimensional tunneling. The primary accomplishment of this past year was to show how one can compute these analytically continued classical trajectories (which are necessarily complex-valued) by direct integration of the classical equations of motion much as one does for ordinary classical trajectories. The first application was made to vibrational excitation in the simple collinear A + EC collision,1 and it was found that one could obtain transition probabilities as small as 10"11 (the smallest available for comparison) with excellent agreement (within a few %) with accurate quantum mechanical values. These results are of great practical importance, for vibrational transitions are typically weak. 2.8 3.0 3.2 3.4 3.6 Application has also been made" to rotational- Ra vibrational transitions for fully three- Fig. 2. Coordinate trajectories for col dimensional collisions of He and H2, again lision energies 0.20 eV (dotted line) and with excellent agreement with quantum 0.02 eV (dash-dot line). For reference, the calculations. dashed line is the "reaction path" and the cross marks the saddle point. Ra and ra Another example of a classically forbidden are the translational and vibrational process is reactive tunneling; this feature coordinates, respectively, for the arrange ment A + BC. (XBL 724-6181) dominates the important threshold region in cases for which there is an activation barrier for the reaction. Complex-valued classical trajectories have been computed for the collinear H + H2 reaction at energies below the classical threshold.3''' Figure 1 shows the ground state reaction probability as a function of collision energy; the circles and crosses are the quantum mechanical results of two different workers, showing excellent agreement with the semiclassical results. (Below 0.21 eV the purely classical cross section is identically zero.) Figure 2 shows the real part of the tunneling trajectory at two different collision energies, illustrating the fact that the sub-threshold dynamics "cuts the comer" of the potential barrier. Semiclassical evaluation of a Boltzmann temperature average shows that this energy region below the classical threshold is the most important for temperatures below 1000°K. 0.10 o.ie E0 (electron volts) 1. W. H. Miller and T. F. George, J. Chem. Phys. 56, 5668 (1972). Fig. 1. Reaction probability for the ground 2. J.TT. Doll and W. H. Miller, J. Chem. state to ground state H + H2 •» H2 + H Phys. 57, 4428 (1972). reaction as a function of relative collision 3. T.~F. George and W. H. Miller, J. Chem. energy. The crosses and circles are the Phys. 56, 5722 (1972). results of different quantum calculations 4. T.~F. George and W. H. Miller, J. Chem. by other • Titers. OCBL 723-6065) Phys. 57, 2458 (1972). -82- 4. SEMICLASSICAL THEORY OF ELECTRONIC classically forbidden collision phenomena, TRANSITIONS IN LOW ENERGY ATOMIC AND since this is the case for which semiclassical MOLECULAR COLLISIONS INVOLVING SEVERAL NUCLEAR theory makes its most important and rather DEGREES OF FREEDOM* unique contribution. William H. Miller and Thomas F. George Further development of the semiclassical theory of electronic ti msitions is needed, The semiclassical theory of electronic and applications to some specific systems transitions in low energy atom-atom collisions are being planned. Of special interest is is to a large extent a solved problem, the the relationship between the intersecting reason being that such transitions have potential energy surfaces and the coupling significant probabilities only if there is between electronic energy with rotation and a crossing, or avoided crossing of potential vibrational energy. Collisions involving curves. This paper shows how electronic electronic transitions are most common when transitions can be treated in a similar one of the colliding species is in an elec manner for cases, such as an atom-diatom tronically excited state, usually a metastable collision, that involve rotation and state that does not decay too fast radiatively. vibrational degrees of freedom in addition Thus many of the applications of this general to translation. theory will deal with quenching of metastable electronic species, such as More specifically, it is shown how one can combine a classical description of all : 3 heavy-particle degrees of freedom (rotation, 0( D) + N -*• 0( P) + N * , vibration, and translation) with a quantum- state description of the electronic degrees of freedom; furthermore, this can be done where N2* denotes vibrational excitation of without the necessity of incorporating any the ground electronic state of N£. dynamical approximations into the theory (other than the semiclassical approximation to the electronic transition]. If certain approximations are made, however, the general 6. 1972 PUBLICATIONS AND REPORTS theory becomes essentially equivalent to the "surface hopping model" used by Tully and William H. Miller and Associates Preston;1 certain features in this model are thus clarified and put on a firmer foundation. Journals The indications are that this theory 1. W. H. Miller, Classical-Limit Green's should provide an accurate description of Function (Fixed-Energy Propagator) and electronic transitions in low energy molecular Classical Quantization of Non-Separable collisions. Systems, J. Chem. Phys. 56, 38 (1972) (LBL-126). s 2. W. H. Miller, Classical Limit of Fredholm Abstracted from J. Chem. Phys. 56, 5637 Theory for Elastic and Inelastic Scattering; (1972). Inability of Phase Space Integrals to Describe Inelastic Transitions, J. Chem. 1. J. C. Tully and R. K. Preston, J. Chem. Phys. 56, 745 (1972)(LBL-172). Phys. 55_, 562 (1971). 3. C. A. Slocomb and H. F. Schaefer, Molecular Autoionization Lifetimes and Cross Sections for Penning Ionization: Numerical 5. RESEARCH PLANS FOR CALENDAR YEAR 1973 Results for He*(ls2s3S) + H(ls2S), J. Chem. Phys. 56_, 1347 (1972) (UCRL-20709). William H. Miller 4. S. M. Hornstein and W. H. Miller, Ways are being explored for simplifying Quantum Corrections (Within the Classical the application of classical S-matrix theory Path Approximation) to the Boltzmann Density to molecular collisions. In particular, ways Matrix, Chem. Phys. Lett. 13, 298 (1972) will be developed to-combine a Monte Carlo (LBL-435). treatment of some degrees of freedom while others are quantized semiclassically. This 5. W. H. Miller and T. F. George, Semi- will greatly simplify the application of the classical Theory of Electronic Transitions in theory to three-dimensional collision Low Energy Atomic and Molecular Collisions systems, both reactive and non-reactive. Involving Several Nuclear Degrees of Freedom, Emphasis will continue to be placed on J. Chem. Phys. 56, 5637 (1972)(LBL-466). -83- 6. IV. H. Miller and T. F. George, Analytic Papers presented Continuation of Classical Mechanics for Classically Forbidden Collisions Processes, 1. T. F. George and W. H. Miller, Classically J. Chem. Phys. 56, 5668 (1972)(LBL-496). Forbidden Collision Processes, Summer Research Conference on Theoretical Chemistry, 7. T. F. George and W. H. Miller, Complex- Boulder, Colorado, June 26-30, 1972. Valued Classical Trajectories for Linear Reactive Collisions of H + H2 Below the 2. W. H. Miller, Semiclassical Theory of Classical Threshold, J. Chem. Phys. 56, 5722 Inelastic and Reactive Scattering, Symposium (1972)(LBL-813). "~ on Bimolecular Reactions as Elementary Processes, Oberwohlfach, Germany, September 8. T. F. George and W. H. Miller, Classical 27-29, 1972. S-Matrix Theory of Reactive Tunneling: LBL reports Linear H + H2 Collisions, J. Chem. Phys. 57, 2458 (1972)(LBL-840). 1. J. D. Doll, T. F. George and W. H. Miller, 9. T. F. George and W. H. Miller, Open- Complex-Valued Classical Trajectories for Channel Projectors for Rearrangement Reactive Tunneling in Three-Dimensional Processes in Molecular Collisions, Phys. Collisions of H and Bz, LBL-1153 (submitted Rev. A, 6, 1885 (1972)(ISL-SZZ). to J. Chem. Phys.). 10. J. D. Doll and W. H. Miller, Classical- 2. W. H. Miller, Improved Classical Path Limit Quantization of Non-Separable Systems: Approximation for the Bolt-man Density Multi-Dimensional WKB Perturbation Theory, Matrix, LBL-1122 (submitted to J. Chem. J. Chem. Phys. 57, 4428 (1972)(LBL-1100). Phys.). 11. J. D. Doll and W. H. Miller, Classical 3. S. A. Augustin, P. K. Pearson, S-Matrix for Vibrational Excitation of H2 H. F. Schaefer and W. H. Miller, Potential Curves and Inelastic Cross Sections for Low by Collision with He in Three Dimensions, + J. Chem. Phys. 57., 5019 (1972)(LBL-U28). Energy Collisions of 0 and He, LBL-1187 (submitted to J. Chem. Phys.). -84- RoVlie J. Myers, Principal Investigator The above equation gives 1. MOLECULAR FIELD CORRECTIONS FOR ELECTRON PARAMAGNETIC RESONANCE SPECTROSCOPY Table I. Molecular Field Corrections for Cu2+ Mag. field Observed Correct g Correct g g value orientation g value value value ZnSe04 (Jindo) (ours) A axis 2.681 2.372 2.172 2.179 C axis 2.491 2.337 2.283 2.278 Y axis (1) 2.489 2.200 2.083 2.098 Y axis (y) 2.850 2.520 2.241 2.260 -85- 2. ELECTRON PARAMAGNETIC RESONANCE SPECTRUM 3+ OF Ti(H2C»6 1 1 Richard C. Wilson and Rollie J. Myers Previous work on Ti(III) has indicated 7.08 that its EPR spectrum could only be detected at very low temperatures. Such observations are consistent with a (3d)1 ion in a coordina tion site with only small distortions from I 6.56 . octahedral. As a result, we were naturally 0%> surprised last year when we observed an EPR •5 spectrum for Ti(III) in aqueous solutions at room tenperature. Since the important species A. 3+ I- in such solutions is an octahedral Ti(H20)0 this observation was quite unexpected. - To confirm the fact that our spectrum was - not due to a complexed or hydrolyzed Ti(III) we used both HC1 and HSO3CF3 to provide vari 5.01 • 1 1 1 able amounts of strong acid. In addition, we 3.410 3.666 3.922 4.IT8 4434 4.690 cooled these acidic solutions until they -1 formed a glass. From this we observed a 1000/T (»K ) simple EPR spectrum characteristic of a ran 3 domly orientated species which follows an Fig. 1. Line width plot for Ti(H20)6 * in axial spin Hamiltonian with g. = 1.98 and water acidified with HC1. Solid points are g = 1.91. n for solution and open points are for solution ± frozen into a glass. The line represents the At room temperature, acidic solutions of least squares fit to a straight line for the Ti(III) give a single broad line about natural logarithm of the line width in gauss 2000 gauss wide. As the temperature is vs. 1000/T (°K-!). (XBL 731-2040 ) lowered, the line width decreases, and at about 228°K the line becomes asymmetric due to the anisotropy of the g tensor. Measure ments of the line width were made in the orbit excited states. This process is known temperature range from 213 to 288°K on as resonant relaxation.1 Relaxation of this various acidic solutions. For each solution, type requires absorption of a lattice vibra an Arrhenius plot of line width versus tempera tion, causing a transition to a real excited ture resulted in a straight line (see pig. 1). state and then relaxation to the ground state Least squares fitting of a straight line to through emission of another lattice vibration. the experimental data gave an activation If resonant relaxation is the predominant energy of 1800*50 cm"*. An unusual feature relaxation process, the line width should be is that no discontinuity occurs at the proportional to exp(-6/kT), where S is the melting point of the solutions. energy separation between the ground state spin doublet and the excited state through Since both gn and gt are close to 2.00, which relaxation is occurring. From our data, we can conclude that the orbital degeneracy we are seeing relaxation through an excited 3 -1 associated with the Ti(H20)0 * must be split state 1800 cm above the ground state. by over 1000 cm"1. Since there are no net Possible sources of lattice vibrations in the lattice distortions to be expected in solu region of 1800 cm"1 are distortion modes of tion, we are forced to conclude that the the octahedral shell of water molecules in Ti(H20)63+ has a Jahn-Teller distortion the complex. Resonant relaxation is also which splits the orbital degeneracy by this frequency independent. Experiments are large amount. If we use the simple formula planned at other frequencies to test whether this is true of the relaxation we are observing. g, = 2.00 - Q . the g values indicate a distortion of about 1 1. C. B. P. Finn, R. Orbach, and W. P. Wolf, 2.000 cm" . Proc. Phys. Soc. 77_, 261 (1961). 'ihis conclusion is consistent with the observed broadening, for what we appear to be seeing is relaxation through one of the spin- -86- 3. PROTON NUCLEAR MAGNETIC RESONANCE ON 10" SOLUTIONS OF TRANSITION-METAL IONS 1 1 1 1 1 1 1 1 Thomas M, Hynes and Rollie J. Myers A number of transition-metal ions have very short electron spin relaxation tires in • solution. These ions cannot be studied by •o*o° " the usual electron paramagnetic resonance 0 techniques, but in some cases the proton NMR of these solutions can depend upon electron 0 spin relaxation. With this aim in mind we have done proton NMR on CofHoO)^2* «md • Ni(H20)62+ in water solutions. n 0 For the species Co(H20)62+ we have measured the transverse spin relaxation tiroes of bulk laD 0 water protons in dilute aqueous solutions of Co(0.04)2 at 60 MHz and 100 MHz and at tem peratures from -35 to +84°C. Low tempera • tures were achieved by 'ising a solution - 10"'- • - which was also 2.46 malar in CaCC104)2. O"1 DM - aim tes been to determine T2M; the transverse n relaxation time of protons in the first ! O '. coordination sphere of the hydrated cobalt ion, over a range of temperatures and fields. d This relaxation arises due to the dipolar coupling of the proton with the electronic a 100 MHz, 5.27 M HCI04 spin.l and its rate is governed by the • • 60 MHz,5-27 M HCIO4 magnitudes of Tj and T2 of the electronic spin. 2 . O 100 MHz, unacidified At temperatures above about 60°C, the • 60 MHz, unacidified transverse relaxation time of the bulk water protons is directly proportional to T^ (T2M region). At lower temperatures this bulk proton relaxation time is also affected by a in) mechanism, in which protons lose phase coherence by jumping between the magnetically different bound and bulk 10 1 11 environments.3 This effect was not observed 2.8 3.2 2.6 4.0 4.4 in earlier work.4'5 At temperatures below about -30°C, bulk proton relaxation begins I03/T ( »K ') to be rate limited by the exchange rate of Fig. 1. The ratio, fy, of coordinated protons protons between the bound and bulk environ to total protons in Ni2+ solutions times T2P, 17 ments. By comparison with 0 work,® the the transverse relaxation tiss of the bulk proton and water exchange rates appear to be protons is plotted vs. the reciprocal of the equal in unacidified solutions. absolute temperature. (XBL 731-2041) Despite the competing effects, values for T2M may be obtained at all temperatures by Similar measurements have been made on using the equations of Swift and Connick,* dilute solutions of Ni (£104)2- Again, the the water exchange rates of Chmelnick and relaxation of protons in the first coordina Fiat ,6 and the proton-cobalt scalar coupling tion sphere of the hydrated nickel ion is constant of Matwiyoff and Darley.l However, dominated by dipolar coupling.7 At tempera determinations of T2M in unacidified solutions tures above room temperature the relaxation are less exact at lower temperatures, 'vhere time of the bulk water protons is directly the Aw and chemical exchange effects become proportional to T^f. At lower temperatures relatively more important. For this reason in unacidified solutions the bulk proton we also made measurements on a solution of relaxation is rate limited by chemical ex Co(H) in 5.27 M HCIOV By increasing the change of protons. This proton exchange rate proton exchange rate, the Au and chemical can be dramatically increased by the presence txchange effects were reduced, and the T2M of acid** and measurements were also made on a region was exte»'"d to somewhat lower tempera solution of Ni(II) in 5.27 M HCIO4 as shown tures. in Fig. 1. This extended tRe T2M region far -87- below 0°C. Curve fitting and the establish and EPR which will be designed to determine ment of relaxation times for both ions awaits the mechanisms for electron spin relaxation. measurements at 220 MHz. Our electron-electron double resonance experiments will be continued so that we can obtain direct evidence of the relative 1. N. A. Matwiyoff and P. E. Darley, J. Phys. importance of dipole-dipole and exchange Chem. 72, 26S9 (1968). interaction for interacting free radicals. 2. R.~E. Connick and D. N. Fiat, J. Chem. In addition, we plan to use our previously Phys. 44, 4103 (1966). developed electrolytic reduction technique 3. T.~T. Swift and R. E. Connick, J. Chem. to do EPR work on some usual radical species Phys. 37, 307 (1962). in solution. 4. R.~Hausser and C. Laukien, Z. Phys. 153, 394 (1959). 5. R. A. Bernheira, T. H. Brown, H. S. Gutowsky, and D. E. Woessner, J. Chem. Phys. 30, 950 (1958). 5. 1972 PUBLICATIONS AND REPORTS 57 A. M. Chmslnick and D. Fiat, J. Chem. Phys. 47, 3986 (1967). Rollie J. Myers and Associates 7. J.~W. Neely and R. E. Connick, J. Amer. Chem. Soc. 94, 3419 (1972). Journals 8. T. J. Swift and G. P. Weinberger, ibid., 90, 2023 (1968). 1. A. Jindo and Rollie J. Myers, Electron Exchange for Hydrated Ions in the a-NiSO,). 6H2O Lattice--Electron Paramagnetic Resonance Study of Hydrated Transition-Ions in 4. RESEARCH PLANS FOR CALENDAR YEAR 1973 ZnSe04.6H2O, J. Mag. Res. 6, 633 (1972). Rollie J. Myers 2. Rollie J. Myers, Electron Exchange for Hydrated Ions in the a-NiS04-6H20 Lattice The molecular field model will be used to Pure and Applied Chem. 32 (1-4), 263 (1972). modify the usual spin Hamiltonian for the presence of interacting ions. As a test case Presented paper for ions with zero-field splittings we will apply this method to Mn2+ in NiSC>4-6H20. We Rollie J. Myers, Electron Exchange for will also apply our method to previously Hydrated Ions in the a-NiS04-6H20 Lattice-- published EPR results done by other workers. The Molecular Field Model and Magnetic Resonance Spectroscopy, Hudson Conference on Further measurements will be made on Magnetic Properties of Transition Metal Ion Ti(H20)63'1' and similar hydrated transition- Complexes,State Univ. of New York, Plattsburgh, metal ions. This work will include both J*fR June 19-21 (1972). -88- Kenneth S. Pitzer, Principal Investigator in which second virial coefficients for pure electrolytes decrease rapidly with increase 1. SOLUTIONS OF ELECTROLYTES in ionic strength in the very dilute region yet become nearly constant at higher con Kenneth S. Pitzer and Guillermo D. Mayorga centrations. This pattern of behavior was known experimentally and seemed puzzling The general objective of this program is heretofore. the prediction of the thermodynamic properties of electrolytes with high accuracy from a By including the appropriate ionic strength minimum number of readily measured parameters. dependence in the second virial coefficients, The equations are selected in the light of a relatively simple system of equations for modern statistical mechanical theories in a the thermodynamic properties of electrolytes manner such that each parameter has general is developed. These equations are shown to physical meaning but need not necessarily be represent experimental behavior to much defined in terms of a detailed molecular higher concentration than any previous sndel. system of comparable complexity. In the first report numerical parameters and comparisons More specifically, it has been shown from with experimental data are given for the an improved statistical treatment based upon osmotic coefficients of several pure the Debye-Huckel ion distribution function electrolytes, for the activity coefficients that the effect of short-range forces between of a series of mixtures HC1-MC1, and for the pairs of ions, i.e., the second virial coef osmotic coefficient for all possible mixtures ficient, should depend also on the ionic of sodium and potassium chlorides and strength. Numerical calculations on more nitrates. The last comparison is shown in realistic models confirm this general effect Fig. 1, where AQ is the difference in the osmotic, coefficient of the mixed electrolyte from the appropriately weighted mean for the pure components at the same ionic strength. The curves for NaCl-KN03 and KCl-NaNO, are predicted from theory and parameters deter mined from the pure electrolytes and mixtures with a common ion, and these curves agree ANoCI-NoNOj satisfactorily with the measured points. VKCI-KNO3 ONaCI-KCI In later work (after the first report) all •KNO, the experimental data for osmotic or activity coefficients of aqueous pure electrolytes at room temperature have been analyzed and the resulting tables of parameters will be published soon. There is good agreement with the general equations, even for 4-1 and 5-1 electrolytes. For 2-2 and 2-3 electrolytes it is necessary to include an additional term, still within the same system of equations, after which agreement is also very good. KCI-NaNOj 2. OTHER TOPICS 0.1 Lawrence S. Bernstein, Janice J. Kim, and Gerald J. Vogt Fig. 1. The difference in osmotic coefficient of mixed electrolytes from pure components at These graduate students have just begun the same ionic strength. The curves for research on topics which are discussed in the next section. NaCl-KN03 and KCl-NaN03 are calculated with parameters from pure electrolytes and mixtures with a common ion. (XBL 725-6295] -89- 3. RESEARCH PLANS FOR CALENDAR YEAR 1973 plausible static deformation. Previous con cepts of pseudorotation and inversion are Kenneth S. Pitzer being generalized from one to three dimensions in an attempt to find a model conforming to a. Solutions of Electrolytes the peculiar properties of XeFg. The relationship of these peculiarities to the The theoretical program will continue with atomic structure of xenon will also be con primary emphasis on mixed aqueous electrolytes. sidered, and this study may lead to con In addition x-ray scattering measurements will sideration of other differences of heavy be made for a few pure solutes with very heavy atoms from the lighter members of the same ions in an effort to provide a direct experi group in the periodic table. mental test of the radial distribution functions calculated theoretically. b. Nuclear Spin Species of Methane 4. 1972 PUBLICATION ANT REPORT The heat capacity of solid methane below Kenneth S. Pitzer 10°K shows anomalous properties presumably associated with conversion of nuclear spin Journal species. This quantity will be measured under conditions providing better control of 1. K. S. Pitzer, Thermodynamic Properties of such spin-species conversion processes in Aqueous Solutions of Bivalent Sulfates, an effort to understand these interesting J. Chem. Soc. Faraday Trans., II, 68, 101 effects. (1972) — c. Anomalous Internal Motions in Molecules LBL reports The molecule XeFD is known to differ 1. Kenneth S. Pitzer, Thermodynamics of greatly from other hexafluorides, which are Electrolytes. I. Theoretical Basis and simple octahedral molecules. But XeFg does General Equations, LBL-846, May 1972 (accepted not display the properties expected for any for publication, J. Phys. Chcm.). -90- E. NUCLEAR SCIENCE Donald R. Olander, Principal Investigator 2. A MODULATED MOLECULAR BEAM STUDY OF THE ENERGY OF SIMPLE GASES SCATTERED FROM 1. LASER VAPORIZATION OF SOLIDS PYROLYTIC GRAPHITE* Robert A. Olstad and Donald R. Olander Wigbert J. Siekhaus, James A. Schwarz and Donald R. Olander The laser-induced vaporization of iron and of a binary solid compound, zirconium The energy of rare gases and several hydride, was studied using a quadrupole mass diatomic gases after scattering from spectrometer. A Q-switch ruby laser with pyrolytic graphite was determined by molecu pulse width of 80 nanoseconds was used to lar beam-phase sensitive detection methods. heat zirconium hydride samples under high The angle of incidence of the molecular beam vacuum to temperatures up to 2500°C. The and the angle of sampling of the reflected mass spectrometer ionizer was located in a molecules were both fixed at 45°. The phase line-of-sight path 40 cm from the target. shifts induced by changing either the tempera By an analysis of the measured mass spectrom ture of the incident gas or the solid were eter signals, the instantaneous rate of utilized to determine an apparent temperature vaporization of hydrogen during the rapid tem of the scattered molecules. perature transient could be determined. The results showed that the apparent Application of an equilibrium model to the temperature of the reflected molecules was vaporization process predicts that pre independent of the temperature of the incident dominantly H2 (rather than H-atoms) should gas. With increasing solid temperature, be emitted. The vaporized H2 should have a the reflected gas temperature increased up Maxwellian velocity distribution characteris to a limiting value which was dependent on tic of the instantaneous surface temperature, the gas and the surface only. Both the basal and the rate of vaporization should be related and prismatic faces of graphite were investi to the H2 vapor pressure by the Langmuir gated. In addition, a prism plane specimen equation. Since the H2 vapor pressure is which had not been subjected to high tempera a function of both surface temperature and ture heat treatment was studied. The apparent surface hydrogen concentration, it was temperature of the reflected beam did not necessary to solve simultaneously the time- depend upon the crystallographic face of the dependent heat conduction and hydrogen graphite but was markedly reduced by high diffusion equations in the heated zirconium temperature annealing. hydride in order to predict the vaporization rates. It was found that H-atom and H2 molecules Abstract of paper published in Surface were emitted at thermal equilibrium with the Science 33, 445 (1972). surface. At low laser power density, the magnitude of the H2 signals agreed with the equilibrium model. At high laser power density, the magnitude of the H2 signals 3. MOLECULAR BEAM SURFACE REACTIONS-WATER was less than predicted, possibly because VAPOR ON GRAPHITE of a diffusion barrier at the surface. The H-atom signals were much larger than pre Rajarama T. Acharya, Donald R. Olander, and dicted, indicating that they were produced Alan Ullman by a non-equilibrium surface process. The modulated molecular beam technique Conventional mode laser pulses (1 msec has been applied to investigate the surface duration) were used to vaporize iron samples. reactions which occur when a beam of H2O The measured mass spectrometer signals were strikes a heated graphite surface. The in agreement with equilibrium calculations amplitudes and phase shifts of the reaction if the flow of molten iron was taken into product signals (CO and H2) have been measured account. as a function of surface temperature, chopping frequency, and beam intensity. -91- The signal amplitudes indicate that the and the middle line represents application of H2O reaction at the annealed prism plane of the law of rectilinear diameters to the pyrolytic graphite becomes large (i.e., viscosities of the two phases. After extrapo reaction probability of > 10"3) only at solid lation of this line by ~ 150°C, the viscosity temperatures above 2000°K. Comparable of cesium at the critical point is found to be reaction probabilities for the 02-graphite 0.53 ± 0.03 mP. reaction were attained at considerably lower temperatures. The CO phase shift in the T 1 T H2O-C reaction approaches a limiting value as the chopping frequency is reduced, which indicates that the reaction is primarily influenced by a diffusional step rather than by a surface chemical reaction. However, the asymptotic value of the phase lag is in between that of a simple diffusional process (45°) and a double diffusion mechanism followed by the O2-C reaction (22 1/2°). The apparent reaction probability of CO is as much as an order of magnitude smaller than that of the other reaction product, H2, which suggests that the surface processes leading to CO production are more strongly diffusion-controlled than those which produce H2. Experiments with variable beam intensity Temperature CO demonstrated that the reaction is first Fig. 1. Viscosity of liquid cesium and order in H2O pressure at the surface. The estimation of the critical viscosity by the H20-C reaction is considerably more complex law of rectilinear diameters. The vapor than the O2-C reaction and the mechanism of viscosity is due to Bonilla. the former has not yet been deduced from the (XBL 731-21S7) data. S. DIFFUSION IN LIQUID URANIUM 4. VISCOSITY OF CESIUM UP TO 1600°C Jack Hovingh and D. R. Olander Han-Chung Tsai and Donald R. Olander The diffusion coefficient and saturation concentration of lanthanum in molten uranium The viscosity of liquid cesium has been were measured by using a capillary tube measured by the oscillating crucible method technique. Activated lanthanum was placed to within 150°C of the critical temperature. on top of a uranium ingot in a beryllia Because of the high vapor pressure of cesium capillary tube. The capillary tube was at temperatures above 1000°C, a massive heated to 1200°C, held at temperature for "pressure-vessel" type of oscillating various diffusion times, and cooled. The crucible was employed. Since the crucibles lanthanum concentration was determined as can only be partially filled with liquid, a function of distance in the uranium from the space above the interface contains cesium the lanthanum uranium interface by counting vapor at high pressure. Consequently, the the photons emitted by the decay of the frictional drag on the crucible inner walls activated lanthanum. is in part due to the viscosity of the vapor, and this effect must be considered in The measured diffusion coefficient of extracting the viscosity of the liquid from lanthanum in molten uranium appears to be the data. Since the viscosity of the liquid at least an order of magnitude less than is < mP, special refractory metal alloy predicted by theory. The measured saturation torsion wires which have been carefully concentration of lanthanum in molten uranium annealed were used to prevent excessive appears to be several times higher than damping of the oscillating motion due to previously reported. The higher apparent internal friction. When corrected for vapor saturation concentration nay be caused by drag, the resulting viscosities of liquid lanthanum seepage into a gap formed between cesium are shown as the upper curve in Fig. 1. the uranium ingot and the beryllia capillary The lower curve shows the vapor viscosity, tube during heating. -92- 6. OXYGEN REDISTRIBUTION IN CERAMIC NUCLEAR ture furnace for producing beams of FUELS* partially dissociated hydrogen has been designed for this experiment. Donald R. Olander Construction of the second molecular beam The presence of temperature gradients as apparatus will be completed. The unique large as 6000°C/cm in ceramic nuclear fuel feature of the new device is the horizontal elements causes extensive redistribution of mounting of the target, which permits the constituents of the fuel. The profile holding of fragile targets such as wafers of the concentration of oxygen in solid of ionic crystals. The system can also be UO2+X which is held in such a temperature used to study reactions of gaseous beams gradient was calculated on the assumption with liquid targets. This apparatus will be that oxygen transport occurs via diffusion used to study the reaction of fluorine of CO, CO2, and 02 along fissures in the molecular beams with single-crystal uranium solid (carbon is an impurity in urania dioxide. which produces CO and CO2 when the fuel is held at high temperatures). Oxygen redistribu The gas phase radiolysis of molecular tion occurs by diffusion of CO2 from the cold oxygen by protons from a Van de Graaff to the hot end, where reduction to CO occurs, accelerator will be completed. followed by back-migration of CO to the cold end. A stationary state is established in The study of lanthanum diffusion in liquid the ternary diffusion system consisting of uranium win continue. In order to obtain CO, CO2, and 02. At each point along the more accurate lanthanum penetration profiles temperature gradient, the equilibrium oxygen than are obtainable by the present method, pressure is equal to that simultaneously the detection system has been redesigned to required by the gas phase reaction CO + permit continuous counting while diffusion 1/2 O2 = CO2 and by the oxygen potential of is in progress. the fuel, which is a function of the oxygen content. The calculations show that initial The analytical studies of thermal gradient ly uniform hyperstoichiometric fuel approaches redistribution in nuclear fuels will be perfect stoichiometry (i.e., UO2.00) at the extended to oxygen migration in mixed cold end and becomes enriched in oxygen at uranium-plutonium oxides. Uranium vapor the hot end. The significance of oxygen transport in mixed hyper-stoichiometric redistribution to the performance of nuclear oxides will also be analyzed. fuel elements is that the fuel composition at the cold outer surface of the fuel rod in large part determines the ability of the cladding to withstand corrosive attack by 8. 1972 PUBLICATIONS AND REPORTS the fuel with which it is in contact. Donald R. Olander and Associates Book chapter Abstract of paper published in J. Nucl. Mater. 44, 116 (1972). 1. D. R. Olander, Technical Basis of the Gas Centrifuge, in Advances in Nuclear Science and Technology,17bl. 6 (19721. PP. 105-174. 7. RESEARCH PLANS FOR CALENDAR YEAR 1973 Journals Donald R. Olander 1. D. R. Olander, A Chemistry-Materials Program in a Nuclear Engineering Curriculum, The modulated beam study of the water Trans. Amer. Nucl. Soc. 15^, 57 (1972). vapor-graphite reaction will be extended by utilizing mixed beams of D2O and H2O in order 2. D. R. Olander, R. H. Jones, W. J. Siekhaus to ascertain the extent of isotope mixing and J. A. Schwarz, Investigation of Gas-Solid in the molecular hydrogen reaction product Reactions by Modulated Molecular Beam Mass and in the unreacted water vapor. The Spectrometry, J. Vac. Sci. and Tech. 9, 1429 accumulated data will be analyzed in order (1972) (LBL-189). to identify likely reaction mechanisms. 3. D. R. Olander, W. Siekhaus, J. Schwarz Upon completion of the H2O-C study, the and R. Jones, Reactions of Modulated molecular beam facility will be used Molecular Beams with Pyrolytic Graphite. to study the reaction of atomic hydrogen Part I - Oxidation of the Basal Plane, J. with pyrolytic graphite. A high tempera Chem. Phys. 57, 408 (1972) (LBL-190). -93- 4. D. R. Olander, W. Siekhaus, J. Schwarz 4. J. A. Schwarz, A Modulated Molecular and R. Jones, Reactions of Modulated Molecu Beam Study of the Energy of Simple Gases lar Beams with Pyrolytic Graphite. Part II- Scattered from Pyrolytic Graphite, Eighth Oxidation of the Prism Plane, J. Chem. Phys. International Symposium on Rarefied Gas s?, 421 [1972) (LBL-191). Dynamics, Stanford, July 1972. 5. W. J. Siekhaus, J. A. Schwarz and 5. D. R. Olander, High Temperature Gas- D. R. Olander, A Modulated Molecular Beam Solid Reactions by Modulated Molecular Beam Study of the Energy of Simple Gases Scattered Mass Spectrometry, FOM Institute of Atomic from Pyrolytic Graphite, Surf. Sci. 33, 445 and Molecular Physics, Amsterdam, Holland, (1972) (LBL-843). October 1972. 6. D. R. Olander, Oxygen Redistribution in U02+x, J- Nucl. Mater. 44, 116 (1972) (LBL-801). LBL reports 7. H. C. Tsai and D. R. Olander, The Viscosity of Molten Uranium Dioxide, J. Nucl. 1. T. R. Acharya and D. R. Olander, The Mater. 4£, 83 (1972) (LBL-809). Rate of Oxidation of the Basal and Prismatic Surfaces of Pyrolytic Graphite in the Papers presented Transition Regime Between Chemical and Diffusional Control, LBL-488, January 1972. 1. D. R. Olander, The Viscosity of Molten Uranium Dioxide, American Nuclear Society 2. D. R. Olander, Oxygen Potential Changes Meeting, Las Vegas,. June 1972. in Mixed Oxide Fuels, LBL-872, June 1972; submitted to Nuclear Technology. 2. D. R. Olander, A Chemistry-Materials Program in a Nuclear Engineering Curriculum, 3. D. R. Olander, The Kinetics of Actinide American Nuclear Society Meeting, Las Vegas, Redistribution and Pore Migration by Vapor June 1972. Migration in Mixed Oxide Fuels, LBL-1131, August 1972; submitted to J. Nucl. Mater. 3. W. J. Siekhaus, A Modulated Molecular Beam Study of the Reaction of Oxygen with the 4. R. A. Olstad, Mass Spectrometric Prismatic Surfaces of Pyrolytic Graphite, Investigation of Laser-Induced Vaporization Eighth International Symposium on Rarefied of Binary Solid Compounds (Ph.D. thesis), Gas Dynamics, Stanford, July 1972. LBL-1177, December 1972. II.Metallurgy -97- A. CRYSTAL IMPERFECTIONS Jack Washburn, Principal Investigator &3S# 1. MECHANISM OF DIPOLE BREAK-UP Jack Washburn Breaking of dislocation dipoles in magnesium oxide into strings of prismatic loops has been followed by repeated electron K. microscope observation of the same areas with intermediate heating of the foil to *\ high temperatures. The break-up mechanism is shown schematical ly in Fig. 1. A wave-like pattern starts from both ends of the dipole in the following way: In an initially uniformly spaced dipole vacancies have a higher chemical potenial •%m at the ends than at any other place due to line tension. Therefore, on annealing, vacancies at the left of the line MN tend to migrate toward the right. This process increases the spacing at 'a' which results in starting a drift of vacancies also from Fig. 2. Breaking of dislocation dipoles, b to a. This further increases the spacing b = 1/2[101], on annealing. (XBB 722-1023) at a and starts to decrease it at b. The reduced spacing at b then starts a flow of vacancies from fat o c. The process repeats A theoretical analysis! based on this toward the middle of the dipole. Thus, a model gave a value of 60,500 cal/mol for wave pattern is set up which eventually the activation energy for pipe diffusion of results in breakup of the dipole into a oxygen ions in MgO. string of loops. As the annealing time increases, the amplitude of the spacing fluctuations increases and the first loop pinches off at 'b', then the next one at 'd' 1. J. Narayan and J. Washburn, Cryst. and so on. See dipoles B, D and E in Fig. 2. Lattice Defects 3, 91 (1972). If the two dislocations in the dipole are not uniformly spaced initially, the effect of non-uniform spacing may cause break-up 2. THE MECHANISM OF DISLOCATION LOOP to start also at interior points. (See COARSENING IN MgO dipole A in Fig. 2.) Jagdish Narayan and Jack Washburn After break-up, wavelength (X) can be measured because it is equal to the spacing When magnesium oxide is plastically of the resulting dislocation loops. deformed (bending) most of the damage is introduced in the form of dislocation dipoles. In a thin sheet bent along the [010] axis, dislocation dipoles are introduced on £101) [101], (101)[101] and on (110)[110], (110) [110] slip systems. If the beading axis is slightly off from [010], dipoles on (011) [011] and (Oil) [011] also appeal'. A brief description of the mechanism of dislocation dipole formation follows: moving screw dislocations acquire jogs of both kinds during motion; this results in formation of edge dislocation dipoles trailing behind. Fig. 1. Schematic illustration of dipole Many of the dipoles are terminated, becoming breaking by pipe diffusion as the time of elongated closed loops by the mechanism annealing is increased. (XBL 725-6220) described by Washburn.1 -98- As the plastically deformed crystals are heated to successively higher annealing 1. J. Washburn, in Electron Microscopy and temperatures, the 'Ipoles start breaking Strengths of Crystals, edited by G. Thomas up into strings of small circular prismatic and J7 Washburn (lnterscience, New York, dislocation loops. First to break up are 1963), Chap. 6. the dipoles of smallest spacing. The break 2. G. W. Groves and A. Kelly, J. Appl. Phys. ing up into circular loops starts at as low 34, 3104 (1963). as 8S0°C, depending upon the separation of 3. J. Narayan and J. Washburn, Phil. Mag. the two edge dislocations comprising the di- 26, 1179 (1972). pole. As the time and temperature of anneal ing is increased, dipoles of larger spacing break up and finally loop coarsening takes place. A Up to now the mechanism of loop coarsening & •> was not clear and was ambiguously attributed to bulk diffusion^ because the thin foils for .2 transmission electron microscopy were obtained from different bulk samples, annealed at dif ferent temperatures for various amounts of time. To investigate the mechanisms of an nealing in more detail, it is necessary to photograph the same area of a single speci *? men after various annealing treatments. Experiments of this type have been carried out with the following results. After the breaking of .dipoles, many pairs of the re 2 sulting coplanar groups of loops, where the > separation was small enough for a strong in teraction, movsd closer together by conserva tive climb or self climb. Self climb of dis location loops differs from the usual non- conservative growth or shrinkage in the sense that the total area of the loop projected 4. • perpendicular to its Burgers vector remains imchanged and climb occurs by migration of at3ms along the easiest path, i.e., the core of the dislocation. 4 Figure 1 shown two pairs of loops during a jUMk&5?A i set of annealing treatments at 1186 and Fig. 1, Self-climb at 1 and 2 at two dif 1250°C. Assuming that the loop migration ferent temperatures. Some of the smaller took place entirely by pipe diffusion along loops have slipped out of the foil. Annealing the circumference of the loops and that the treatments: A-*B, 2438 sec at 1186°C; B->C, driving force was only the loop-loop interac 2392 sec at 1250°C; C->D, 1570 sec at 1250°C; tion energy, an expression for the rate of D*E, 1202 sec at 1250°C; E*F, 2369 sec at loop migration was derived.3 Quantitative 1250°C. (XBB 721-117) measurements of the rates of migration of seven loop pairs lead to the following conclusions: 3. PRECIPITATION OF PHOSPHOROUS IN ION 1) The activation energy forppipe diffusion IMPLANTED SILICON and pre-exponential factor Do a™from measurements of rate of self-climb of Krishna Seshan and Jack Washburn coplanar pairs of loops, b = 1/2 <101 ) , are 60,300*3500 cal/mol and (7.5±4.3)xl0-:'-8 cm4/ Ion implantation in silicon is accompanied sec, respectively. by damage along the path of each implanted 2) Self-climb is the primary mechanism ion which above a critical number of ions of dislocation loop coarsening during per cm? of the order of 101* results in an annealing of plastically deformed magnesium amorphous layer. On subsequent annealing oxide up to 12S0°C. recrystallization of this layer occurs and -99- Fig. 1. 'feak beam dark field image of a p- type silicon sample irradiated with 2 x 1014 p ions/cm2. The foil is oriented for a two-beam case. The linear defects lie along the <110 > directions. (XBB 722-750] interstitials and vacancies aggregate into Davidson and Booker.* Dark field and weak defects large enough to be detected by beam dark field observations suggest that the transmission electron microscopy. The defects cannot be simple dislocation dipoies. present experiments are part of a continuing The type of bl?£k-white contrast which appears study of the nature and effects of these for rods that lie parallel to the foil surface defects. is also similar to that which is observed for the rod-shaped precipitates in Al-Mg^Si.5 Figure 1 shows a specimen that was implanted with 2 * 1014 phosphorous ions/cm2 It is found that rod-like defects are then heated to 800°C for 1/2 hour. In absent from n-type silicon of the same dose addition to dislocation loops there are and heat treatment. This indicates that the numerous rod-shaped defects that lie along small amount of boron, already present, has an <110 > directions. Similar rod-shaped important influence on the formation of linear defects have been observed in boron implanted defects. However, the number of boron atoms silicon**- and in silicon implanted with in the 1 micron thick implantation layer is phosphorous at 200°C.3 These rod defects only about 10lz per cm2. This alone is in have previously been reported to be absent sufficient to cause such a high density of when phosphorous is implanted at 20°C.1>3 defects. Therefore, it was concluded that these rod defects also involve the precipita The diffraction contrast effects exhibited tion of phosphorous. by the rods are similar to those described by -100- 1. R. W. Bicknell, European Conference on Ion Implantation, IPPS/IEE. September 1970. 2. L. T. Chadderton and F. H. Eisen, Rad. Eff. 7, 129 (1971). 3. M7 Tamura, T. Ikeda and Tokuyana, Proc. Second International Conference on Ion Implantation in Semiconductors, edited by I. Ruge and J. Graul, (Springer-Verlag, 1971), p. 46. 4. S. M. Davidson and G. R. Booker, Rad. Eff. 6, 83 C1970). 5. W7 L. Bell and G. Thomas, Proc. 6th International Congress for Electron Micros copy, Kyoto (1966). 4. MEMORY STATE FILAMENTS IN As-Te-I AMORPHOUS SEMICONDUCTORS Jon Roberts This study investigates the mechanism of growth and the crystalline character of the conductive filaments produced electrical ly in As.53Te.4-5I.04 amorphous alloy. The filaments begin growth at the positive electrical contact after the material has switched into the conductive state, and grow at a rate proportional to the square root of the duration of power application. The level of applied current has no signifi cant effect on the growth rate, so it is Fig. 1. Annealing sequence showing a cluster concluded that the driving force for diffusion emerging on a (100) plane. (XBB 731-277) is the electric field. The filaments are composed of tellurium proton bombardment of iridium.1 These spots crystals, as determined by electron beam have been attributed to interstitials microprobe studies. A method of extracting migrating to the surface. An additional the filaments for electron diffraction observation in this study is the appearance studies has been developed, and the results of large clusters of bright spots, containing show that the filaments are often single from 3 to as many as 15 spots. One example crystals. Continued studies will determine of a cluster is shown in Fig. 1. The the preferred growth orientation. clusters have been observed to pop up on the surface over the same temperature interval as the single and double spots. After emerging, the clusters tend to be 5. FIELD ION MICROSCOPE OBSERVATIONS OF LOW stable with further annpaling, i.e., no spots TEMPERATURE ION BOMBARDMENT DAMAGE IN disappear when the temperature is increased. IRIDIUM This is taken as evidence that the clusters are not due to surface migration of impurity Kaj Stolt atoms from the shank of the tip. This is also supported by the fact that no clusters Well-annealed Ir specimens are beinr have been observed during annealing of irradiated at 10°K with 30 kV argon ids unirradiated specimen?. Preliminary investi in situ in the atom probe FtM described in gations of the behavior of the clusters the previous report. After bombardment the with increasing ,1 voltage reveal that the surface damage is removed by field spots can be removed one at a time by pulsed evaporation of a few surface layers, after field evaporation. This seems to indicate which the specimen is slowly annealed up to that the spots are due to individual atoms 70°K. During annealing single and double in metastable positions on the surface. bright spots are seen to pop up on the surface, in the temperature interval 15-70°K, At present an attempt to clarify whether as has been observed previously following impurity atoms play a role in formation of -101- the clusters is being undertaken by analyzing 7. 1972 PUBLICATIONS AND REPORTS the clusters with the atom probe. Investiga tion of the volume of material immediately Jack Washburn and Associates beneath a cluster by careful field evaporation should also shed more light on the orgin of Journals and book the clusters. 1. J. Narayan and J. Washburn, Direct Observation of Pipe Diffusion in Magnesium Oxide, Scripta Met. 6, 259-262 (1972) 1. P. Petroff and J. Washburn, A Field Ion Microscope Investigation of 10 MeV Proton (LBL-476). Damage in Iridium, Phys. Stat. Sol. 32, 527 (1969). 2. J. Narayan and J. Washburn, Direct Evidence of Presence of Both Interstitial and Vacancy Dislocation Loops in Plastically Deformed and Subsequently Annealed Magnesium 6. RESEARCH PLANS FOR CALENDAR YEAR 1973 Oxide, Scripta Met. 6, 263-266 (1972) (LBL-47S). Jack Washburn 3. J. Washburn, Annealing of Loops and Tetrahedra, in Radiation-Induced Voids in 1. Electron microscope studies of ion- Metals (Proc. o± the 1971 Internatl. Conf. implanted silicon: Effects of annealing Albany, N.Y., June 9-11, 1971), edited by treatments and ion implantation dose on de J. W. Corbett anri L. C. Ianniello, pp. 647- fects in the implanted layer will continue to 662. be investigated for phosphorous-implanted silicon, and the work will be expanded to study arsenic and boron-implanted silicon. 4. E. Nes and J. Washburn, Precipitate Colonies in Silicon, J. Appl. Phys. 43, 2005 — 2. Field ion microscopy will be applied (1972)(UCRL-20354 Supp.). to a study of the effect of surface layer atoms on work function. Ni4W in the ordered 5. J. Narayan and J. Washburn, Pipe Diffusion state permits surface layers consisting in Magnesium Oxide, Crystal Lattice Defects entirely of Ni atoms or of Ni + W to be 3, 31-98 (1972)(LBL-492). exposed by field evaporation in certain regions of the field ion microscope tips. 6. K. Seshan and J. Washburn, On Precipita Work function measurements iri.ll be compared tion of Phosphorous in Ion Implanted Silicon, for the two kinds of surfaces. Radiation Effects 1£, 267 (1972)(LBL-493). 3. Radiation damage produced by election 7. J. Narayan and J. Washburn, Self-climb bombardment of copper and copper aluminum of Dislocation Loops in Magnesium Oxide, alloys in the 650 kV election microscope Phil. Mag. 26, No. 5, 1179 (1972) (UCRL-2023S). will be used in a continuing study of dis location climb phenomena. In particular the 8. J. Narayan and J. Washburn, On the effect of stacking fault energy on climb due Stability of Dislocation Loops Near a Free to capture of interstitial atoms produced Surface, J. Appl. Phys., 43 (12), 4862 (1972) by electron irradiation will be investigated. (LBL-819). "~ To make these experiments practical it is planned to modify the colran of the high 9. J. Narayan and J. Washburn, Pipe Diffusion voltage microscope to permit use of side in Magnesium Oxide, Crystal Lattice Defects entry stages. 3, 91 (1972) (LBL-492). 4. The possibility of producing a low 10. J. Narayan, On the Termination of Dis cost photovoltaic cell based on a sputtered location Dipoles, Phys. Stat. Sol. (b) 54, thin film of ZnS will continue to be Kl (1972) (LBL-436). investigated along with a continuing study of the mechanism of operation of the cuprous sulphia<./cadmium sulfide hetrojunction solar cell. Paper presented 5. Memory filaments in amorphous As-Te-I semiconducting glasses have been characterized 1. J. Washburn, Direct Observations of as whisker-like crystals of Te. However, Dislocation Climb, presented at J. Dorn work still remains to be done before an Memorial Symposium, AIME. Cleveland, Ohio, understanding of their mechanism of growth Oct. 16-19, 1972 (LBL-866). will be achieved. -102- LBL reports and Interstitial Loops in Ion Implanted Crystals, LBL-11Q3, August 1972; submitted 1. J. Narayan and J. Washburn, On the to Physica Status Solidi (A). Mschanism of Dislocation Loop Coarsening in MgO (Conservative Climb), LBL-825, March 4. K. Seshan, W. L. Bell, and J. Washburn, 1972; submitted to J. Am. Ceram. Soc. A Rapid Method for Distinguishing Vacancy and Interstitial Loops in Ion Implanted 2. J. Narayan and J. Washburn, Electron Crystals, LBL-1103, August 1972; submitted Microscopic Studies of the Micro-Erosion in to Physica Status Solidi (A). Brittle Solids, LBL-899, June 1972; sub mitted to Wear. 5. J. Naxayan and J. Washburn, Self Diffusion in Magnesium Oxide, Acta Met., in press; 3. K. Seshan, W. L. Bell, and J. Washburn, LBL-473, Feb. 1972. A Rapid Method for Distinguishing Vacancy -103- B. THEORETICAL METALLURGY John W. Movrie, Jr., Principal Investigator J. W. Morris, Jr., J. Chem. Phys. 57, 769 — 1. THERMODYNAMICS AND PHASE STABILITY (1972) J. W. Morris, Jr. and P. L. DeBruyn, LBL-4S4 a. Thermodynamics of Inhoroogeneous Systems (J. Colloid Interface Science (in press)). John W. Morris, Jr. b. Bonding and Phase Stability Recent research* into the nature of Charles W. Krause and John W. Morris, Jr. continuous two-phase interfaces in systems obeying the thermodynamics of one-component The existence of various dose-packed fluids led to the intriguing result that phases provides evidence that relatively material near the interface is necessarily long-range interatomic forces are acting in subject to a non-classical stress system a metal. For example, if it is assumed that which includes "multipolar" stresses of a central two-body potential determines the tensor order greater than 2. It was shown crystal structure of a metal, then the range that despite the mechanical complexity of of the potential must extend out to at least the interface Gibbs1 classical thermodynamics third nearest neighbors in order for a of surfaces follows naturally when the Gibbs crystal to exist in either the hep or the construction is used. However, the multi- fee phase. polarity of the stress within the interface did raise questions concerning the proper This work treats the stability of close- definition of the "surface stress" and the packed structures employing a relatively validity of the claim, based on oversimple long-ranged interatomic potential which mechanical models, that the "surface stress" arises from the nearly free electron model of an interface in a simple fluid is equal of a metal treated to second order in to its thermodynamic surface tension. perturbation theory. The potential exhibits the Friedel oscillations, These questions were investigated by using the second-order gradient model, developed by Van der Waals and by Cahn and Hilliard, V(r) = V CI) in which the free energy density at a point (2Kpr)J within the fluid is assumed to depend on the local temperature, the local mass density, and the local values of the first two and converges to the interatomic potential gradients of the mass density. The stress derived from second-order perturbation theory system within an interface in such a fluid at large values of r. Use of the dimension- less quantity V(r)/V0 allows the relative includes non-zero stresses of tensor order stability of the various phases to be dis 2, 3, and 4. Nonetheless a scalar thermo cussed as a function of a single parameter, dynamic pressure may be defined within the Z, the valence of the metal. This real interface, and the surface tension is given space approach to phase stability is easily by the surface excess of this pressure. visualized and, as formulated here, can v easily be applied to any close-packed The results of this investigation ' e the polytype. following. Let the "surface stress" be defined, as is customary, from the incremental The presert work also clarifies the work done in an infinitesimal stretching of distinction between stability against the surface which preserves its shape. If faulting and stability against phase trans the surface is initially flat this "surface formation. There exist values of Z for stress" is identical to the surface tension. which both the hep and fee phases are stable Only the second-order stress does work during against faulting, but one of the two is a homogeneous stretching; the multipolar unstable against phase transformation. For stresses maintain the equilibrium density other values of Z, neither the hep nor the variation normal to the plane of the surface. fee phase is stable against faulting, and If the surface is initially spherical the a more complex close-packed phase appears. multipolar stresses do non-negligible work during a homogeneous stretching, but the "surface stress" may still be shown to be c. Microstructures equal to the surface tension to within terms of order R"2, where R is the radius of Kenton L. Hanson and John W. Morris. •';. curvature. During the past year we inaugurated -104- research using numerical modeling to simulate allows the mean and variance of the velocity simple nucleation and growth transformations of glide to be obtained analytically from in solids. The research addresses two a properly phrased distribution of the central problems in physical metallurgy: the "strengths" of the obstacle configurations problem of predicting the microstructure encountered by the dislocation as it moves resulting from a given transformation and the through the array. While this distribution converse problem of inferring transformation cannot yet be computed analytically, we are behavior from the resulting n&cTOStructure. now able to predict glide behavior over a wide range of experimental conditions from Our initial subject was the microstructure a single numerical "e:tperrjnent." of sections through a body transformed via random nucleation and isotropic growth. The 2) In contrast to a frequent assumption, code we employ constructs a cube with periodic the velocity of gli-'D obeys an equation of boundary conditions, and nucleates grains at the simple Arrenhim form only in the limits random sites within the cube at times deter of extreme temperature, stress, or obstacle mined from a specified probability of strength. In general the effective activa nucleation. The grains grow isotropically at tion energy for glide is a function of a specified rats. The code then plots the temperature, stress, and obstacle strength. microttructures observed on a specified set Its variation with temperature may be related of plane sections through the cube. The code to the "entropy" of the distribution of was written and used to generate hypothetical obstacle strengths encountered as the microstructures under a variety of conditions. dislocation glides through the array. The topological properties of these micro- structures are under investigation. An 3) Dislocation glide through a random intriguing initial observation is the large array of obstacles is jerky, as has been size variation and irregular, often acicular observed experimentally. In contrast to shapes of the sections of grains resulting frequent assumption, however, the apparent from random nucleation and isotropic growth. "jerkiness" depends strongly on the tempera In conventional metallography these features ture as well as the applied stress. are often cited as evidence for heterogeneous nucleation sites or preferred growth direc Recently our numerical code has been tions. throughly revised. The new form allows easy variation of array size and boundary con ditions , and employs a new search algorithm which ensures that any dislocation "looping" 2. MECHANICAL PROPERTIES will be accurately reproduced. The nev code also allows independent variation of the a. Kinetics of Dislocation Glide properties of the obstacles; hence arrays of unlike obstacles can be studied. Corrolary Dale H. Klahn, Kenton L. Hanson, Choi K. Syn, research on dislocation glide through an and John W. Morris, Jr. array of rod-like precipitates shows that the precipitate field may be regarded as a The mechanical behavior of a crystalline mixed array of these obstacle types. Force- solid is often influenced by dislocation displacement curves have been derived for motion through a field of obstacles, for these types, allowing numerical simulation example, forest dislocations, solute atoms, of glide through an array of small rod-like or small precipitates, which are dispersed particles. in a more or less random fashion through the crystal. The problem of predicting the b. Dislocation-Solute Interactions rate of this dislocation motion at finite temperature is formidable. We have been Choi K. Syn and John W. Morris, Jr. engaged in a detailed study of one of the simplest problems of this type: the thermally The edge dislocation-substitutional activated glide of a dislocation, idealized solute atom interaction in a hexagonal Zn as a line of constant tension, through a crystal containing a solute addition of Cd random array of immobile point obstacles or Ag has been estimated and the xorce- under constant applied stress. The study displacement diagram for rigid glide of the involves theory supplemented by computer dislocation with respect to its solute simulation of the dislocation glide. Results atmosphere has f.een computed. Elastic obtained during the past year include the interaction due to both "size" and "modulus" following. effects was computed; other sources of interaction were neglected. The anisotropy 1] The statistics of thermally activated of the dislocation stress field, the lattice glide have been brought into a form which distortion by solute atoms, and the elastic -105- constants of the lattice and solute atoms tion of Harper-Dorn behavior in the metals were incorporated into the calculations. tested was determined from the linear stress The breakdown of linear elasticity within exponent, the discrepancy between the the dislocation core was artificially removed experimental values of the creep rate and using a modification parameter suggested by that obtainable from various mechanisms of Dorn and Stefansky. The vibrational entropy viscous deformation, and the close correspond effect on the configuration of solute atoms ence between the experimental data and the along the dislocation line was also taken earlier results on aluminum and aluminum-3% into consideration. magnesium. The experimental results on aluminum reproduce the data of Harper and The following results were obtained. Dom almost exactly, despite a significant difference in the sample and load geometries. 1) The interaction energy contours due to the elastic "size" effect are markedly The experimental evidence of Harper-Dorn different in shape from those obtained in creep in lead and tin along with data the isotropic approximation. The solute- indicating the occurrence of the same creep dislocation binding energies due to the behavior in other materials suggests that "size" effect are much higher (0.81 eV for Harper-Dorn creep might be a general mechanism Ag and 0 39 eV for Cd) than those obtained for creep at high temperature and low stress. from isotropic calculations (0.16 eV for Ag While the details of the process are unknown, and 0.22 eV fcr Cd). available mechanical and microstructural data may be interpreted as due to climb of 2) When the elastic "modulus" effect is "saturated" dislocations. superimposed, the interaction energy contours are slightly perturbed within one Burgers The simple Harper-Dorn creep behavior vector of the dislocation core. However, the disappears when the applied stress exceeds contours at greater distances are not modified a critical value or when sample grain size seriously. becomes sufficiently small. In the first case, the rate controlling mechanism changes 3) The F-X diagram for Ag and Cd due to to climb behavior, while in the second case elastic interaction contrast strongly. The Nabarrc-Herring mechanism dominates. F-X diagram for Ag has two maxima, while that foT Cd has one maximum and is similar d. Alloy Design in s'.ape to that obtained assuming isotropy. Sungho Jin and John W. Morris, Jr. 4) Including Dorn-Stefansky core modifica tion lowers the maximum of the F-X diagram During the past year our effort under by a factor of 20% for Ag and by one order this program focused on the development of of magnitude for Cd. tough ferritic alloys for use at cryogenic temperatures. This program is conducted 5) Including the effect of vibrational jointly with Prof. V. F. Zackay and entropy lowers the maximum by 50% for Cd, associates. Accomplishments are reported in but does not sensibly affect the maximum the section on High Field Superconductivity. for Ag. e. Mechanical Testing Techniques for Use c. High Temperature Creep at Cryogenic Temperature Farghalli A. Mohamed, K. Linga Murty and Sungho Jin and John W. Morris, Jr. John W. Morris, Jr. Our interest in cryogenic alloys has High-temperature creep has been studied created a need for simple inexpensive tech in aluminum, lead, and tin in the stress niques for obtaining good mechanical property range of 10~6 - 10"4 G(G is the shear data at cryogenic temperature. During the modulus) using double-shear type specimens past year a new method of Charpy impact at temperatures near the melting point. The testing at temperatures to 5°K was devised. experimental data reveal that at x < 10"= G In this technique a 51 mm Charpy V-notch the dominant creep mechanism changes' from specimen is wrapped in styrofoam and placed climb mechanism to Harper-Dorn creep, a type inside a thin-walled lucite box. Liquid of viscous deformation reported in pure helium is then transferred to the box at aluminum .''ud aluminum-3% magnesium at high 2.3 psi dewar pressure and a flow rate of temperature and low stress. It is character ~ 2 liters/min, using a vacuum-insulated ized by a creep rate which is linearly transfer line. A steady-state temperature dependent on the applied stress and apparently of 5-6°K is quickly attained and is maintained independent of grain size. The identifica constant as long as the helium is being -106- transferred. The measured impact energy is of the results obtained from the numerical calibrated by subtracting the net energy experiments. absorbed by the box, measured to be less than 2 ft lb. Comparison impact tests on Research on the dislocation-solute boxed and standard samples at liquid nitrogen interaction will focus on dislocation break temperature confirm the accuracy of the values away through kink nucleation, and on proper obtained. Tests may be conducted over a range computation of the solute atmosphere under of temperatures by varying the helium pres near saturation conditions. sures and flow rate. 4. 1972 PUBLICATIONS AND REPORTS 3. RESEARCH PLANS FOR CALENDAR YEAR 1973 John W. Morris and Associates John W. Morris, Jr. Journals a. Thermodynamics and Phase Stability *1. J. W. Morris, Jr., The Two-Phase Fluid Our principal research emphasis during the Interface at Equilibrium: A Continuum Model, coming year will be on the use of pseudo- J. Chem. Phys. 57., 769 (1972). potential techniques to treat phase stability in alloys. Two approaches to the problem will 2. K. L. Murty, F. L. Mohamed, and J. E. be investigated. First, using simple model Dom, Viscous Glide, Dislocation Climb, and pseudo-potentials it is possible to construct Newtonian Viscous Deformation Mechanism of prototype phase diagrams for idealized alloy High Temperature Creep in Al-3 Mg, Acta Met. systems. While the assumptions necessary in 20, 1009 (1972)(LBL-442). this computation preclude a quantitative prediction of the phase diagram of a real 3. C. Lin, S. Lau, and J. E. Dorn, The alloy, the detailed study of a simple model Plastic Deformation Behavior of Mo Single system should help to clarify the sources of Crystals Under Compression, Phvs. Status phase stability in real materials of metal Solidi A 11, 645 (1972). lurgical interest. Second, given recent advances in pseudo-potential technique, it 4. P. G. McCormick and K. L. Murty, On the may now be possible to quantitatively treat Effect of Vacancy Sinks and Sources on dilute solid solutions in simple alloys. Key Serrated Yielding Due to Solute Locking, current problems include proper treatment ScriptaMet. 6, 225 (1972)(LBL-194 Rev.). of the solute size effect and the perturbation due to difference in the valence of the alloy 5. F. A. Mohamed, K. Linga Murty, and species. We hope to overcome these problems J, E. Com, On the Kinetics of Ordering in and compute thermodynamic data for the simple CU3AU, Mat. Sci. Eng. 9, 183 (1972) (UCRL- alloy systems in reasonable agreement with 19121 Rev.). experiment. 6. K. Linga Murty and J. E. Dorn, On the b. Mechanical Properties Non-Equilibrium Factor for Nucleation, J. Phys. Chem. Solids, Tech. Note, 33 (3), Our principal research emphasis during 757 (1972)(UCRL-20578). the coming year will continue to be on the theory of dislocation glide through distribu Papers presented tions of obstacles. We plan three complimen tary investigations. First, we hope to 1. D. H. KLahn and J. W. Morris, Jr., complete the theory of thermally activated Dislocation Glide Through a Random Array of glide of an idealized dislocation through a Obstacles, IMS Spring Meeting, Boston, random array of like point obstacles by May 1972. obtaining a theoretical solution for the relevant properties of the obstacle con 2. C. Liu, J. W. Morris, Jr., and J. E. figurations encountered as the dislocation Dorn, Effect of Re on Anisotropic Slip in glides through the array. Second, we plan Single Mo Crystals,TM> Spring Meeting, to use the revised numerical code to study Boston, May 1972. more realistic obstacle models, including distributions of obstacle strengths and 3. K. L. Murty, F. A. Mohamed, partially ordered configurations. Third, we J. W. Morris, Jr., and J. E. Dom, High plan to resurvey the available experimental Temperature Creep Mechanisms in Al-3Mg and data on thermally activated glide in light Other Metals, IMS Spring Meeting, Boston, May 1972. -107- 4. K. L. Murty, M. L. Vaidya, J. W. Morris, 3. C. Liu, J. W. Morris, Jr., and J. E. Jr., and J. E. Dorn, Deformation Mechanisms Dorn, Effect of Rhenium Additions on in Superplastic Zn-22A1 Eutectoid, IMS Anisotropic Slip in Molybdenum Single Spring Meeting, Boston, May 1972. Crystals, LBL-1148, November 1972 (submitted for publication). 5. D. H. Klahn and J. W. Morris, Jr., Characteristic of Dislocation Glide Through 4. M. L. Vaidya, K. L. Muity, and a Random Array of Obstacles, John E. Dorn J. E. Dorn, High Temperature Deformation Memorial Symposium, Cleveland, October 1972 Mechanisms in Superplastic Zn-22A1 eutectoid, (proceedings in press); LBL-1168, November LBL-459, February 1972. 1972. 6. F. A. Mohamed, K. L. Murty, and J. V,', 6. K. E. Amin, K. Linga Murty, and John E. Morris, Jr., Harper-Dorn Creep in Metals at Dorn, Parametrization of Low Temperature High Temperature, John E. Dom Memorial Deformation Characteristics in Single Symposium, Cleveland, October 1972 (proceed Crystals of Molybdenum, J. Appl. Fhys., ings in press); LBL-1158, September 1972. in press (LBL-407 Rev.). LBL reports 7. J. W. Morris, Jr., and P. L. de Bruyn, The Two-Phase Interface in a Simple One- i. D. H. Klahn, Dislocation Modeling of Component Fluid, J. Colloid Interface Sci., Strain-Rate Phenomena (Ph.D. thesis), in press (LBL-454). LBL-800, March 1972. 8. F. A. Mohamed, K. L. Murty, and J. W. 2. F. A. Mohamed, High Temperature Creep in Morris, Jr., Harper-Dorn Creep in Al, Pb, Metals at Intermediate and Low Stresses and Sn, Met. Trans., in press (LBL-831, (Ph.D. thesis), LBL-1183, November 1972. May 1972). -108- C. RELATION OF MICROSTRUCTURE AND PROPERTIES Gareth Thonae, Principal Investigator Introduction. One of the objectives of research relating structure and properties is to be able to design economically, alloys with desirable structures for particular applications and properties. Over the past eight years we have been concerned with two such examples: a) the control of transformation sub structure in martensite and bainites in structural steels leading to strong, tough economical steels b) the development of high strength dis persion strengthened systems without involving plastic deformation. Ihese programs have involved detailed and systematic studies of a series of alloys by quantitative electron microscopy. Success has been obtained iji both these areas; in a) it has been shown that toughness can be achieved in relatively high carbon steels (~ 0.25% C) provided twinned martensite and inter-martensite or bainite carbide films are avoided. In b) the martensite trans formation is utilized to produce dislocation- particle interactions and thus dispersion Fig. 1. Typical as-quenched microstructures strengthening in prior dispersed austenite. of 0.35 wU C alloys. Alloy 0435 contains The latter is achieved by, e.g., ausaging a mixed structure of dislocated laths fa) and or spinodal transformation. twinned plates (b). Alloy 1235 also has a mixed structure, shown in (c) and (d), Work is continuing in understanding phase although the twin density in this alloy is transformations from a fundamental view point, much higher than that of alloy 0435. e.g., studies of martensite nucleation in (XBB 7211-5622) single crystals, bainitic transformations, spinodal and ordering transitions. The abstracts following are representative In the Fe-Cr-C steels studied of the current programs. (4-12 wtS Cr; 0.16-0.35 wt% C), the as- quenched martensitic structure as well as carbide morphologies produced by tempering at 200, 400, and 600°C were correlated with tensile and fracture (K\c) properties. In 1. STBUCIUHE, STRENGTH AND TOUGHNESS OF this paper we present new results concerning Fe-Cr-C MARTENSITIC STEELS* the strength and toughness of the untempered martensites. For a given carbon level, the John A. McMahon and Gareth Thomas volume fraction of martensite containing twins increased with chromium concentration, This research is part of a continuing demonstrating that chromium induces trans program involving a systematic study of the formation twinning in martensite (Fig. 1). composition dependence of martensitic Ms temperature measurements showed that structure as revealed by electron microscopy increasing the chromium content also lowers (dislocated or twinned, carbide type, and the Ms temperature. morphology) and its relation to strength and touguvss. The overall objective is the The mechanical properties of the as- improvement in design of structural steels quenched and tempered structures were deter by simplifying composition and heat treatment mined by using tensile, compressive, and to minimize costs. plane strain fracture toughness tests. It -109- taining 0.35 wtfc carbon, raising the chromium ALLOT content from 4 to 12 wt% caused the percentage HO. o 0417 of twinned martensite to increase from a 0817 a 1217 approximately 20% to greater than 3D% fas • 0435 A 1235 estimated from electron micrographs), and the plane strain fracture toughness to drop from 75 to 25 ksi-inV2, The high chromium alloy also had a higher yield strength: 300 ksi vs 240 ksi for the low chromium alloy. However, when tempered to an equiv alent yield strength, the 12 wt? chromium alloy with a high twin density continued to have much poorer fracture toughness compared to the 4 wtt chromium alloy with a lower twin density (see Fig. 2). Thus it has been established that with the carbon content held constant, there is a direct correlation be- tweenthe transformation twin density, chromi um content (and Ms), and strength and tough ness in Fe-Cr-C steels. 260 Z80 300 One of the results of this systematic Fig. 2. Yield strength versus fracture study of the microstructure and mechanical toughness for all alloys under various properties of Fe-Cr-C steels has been the tempering conditions, giving a strength/ discovery of an alloy with mechanical proper toughness comparison of dislocated and ties comparable to those of 18 Ni maraging twinned martensites. (XBL 7211-71S9) steels (Fig. 3) and which could be produced at a significantly lower cost. We appreciate the continued support of Dr Frank Hultgren, Republic Steel Corporation, who kindly provided all the alloys. Based on LBL-1181 (J. A. McMahon, M.S. thesis). 2. MARTENSITIC TRANSFORMATION AND STRENGTHENING OF A SPINOOAL Fe/Ni/Cu ALLOT Gareth Thomas and C. Vercaemer1' A 61 at.Fe/20 at.Ni/19 at.Cu alloy has been designed which can be subsequently transformed to martensite following spinodal heat treatment (aging at 850°C). Morpho 220 240 260 260 300 logical changes were followed by transmis ULTIMATE TENSILE STRENGTH, KSI sion electron microscopy Figs. 1 and 2. Fig. 3. Comparison of ultimate tensile Ms temperatures were estimated by an elec strength and fracture toughness of alloy trical resistivity technique and changes 043S with the strength/toughness properties in microhardness were measured, 'the Ms of several ultra-high-strength commercial temperatures increased rapidly, then were steels. (XBL 7211-7161) unchanged as particle coarsening started and wavelength A increased. The hardness first increased rapidly, then age-softened was found that &r the same carbon level, an as X increased. increase in the amount of transformation twinning due to an increased chromium con Although the specimens were brittle, centration has a dramatic and deleterious the considerable increases in hardness effect on the fracture toughness of the as- achieved by martensitic transformation quenched steels. In two of the steels con compared to the hardness of purely spinodel -110- ,1 a Fig. 1. Transmission electron micrographs in bright field, (a} As quenched: >, -360A, and aged at 850°C for (b) 3 min, X -205A; (c) 9 min, A -360A; (d) 2 hr, \ -890A. (XBB 7112-5991) Fig. 2. Transmission electron micrographs showing martensite in the 70 hr aged alloy; (a) bright field, (b) dark field of austenite spot reversing contrast. M represents the martensitic area and A the austenitic area. (XBB 7112-5993) -Ill- * Wove length o Temperolure „< 1500 - 0 £1000 \~ - -60 -100 £ 500 i/y -150 i ,, i I I Aging Time) i Fig. 3. Variation of hardness with aging time for samples quenched in water and in liquid nitrogen. Aging temperature 850°C. Livak and Thomas' results have been repre sented for comparison (aged at 625°C compared to 850°C l-i the present alloy). (XBL 7112-2204) alloys (Fig. 3) indicates that this dou ble phase transformation method may be promising for improving the strength, and could replace ausforming in certain alloys. LBL-447, Met. Trans, 3_, pp 2501-2506 (1972). Now returned to Departement Science des Fig. 1. Dark field electron micrographs of Materiaux, Uhiversite de Strasbourg, France. {112} type twinned iron-1.54 nitrogen martensite tempered 1 hr at 200°C showing a" lamellae on (001)a< habit planes in both parent (a) and twin (b) lattices. In (O 3. THE TEMPERING OF MARTENSITE IN AN IRON the operating reflections are the combined 1.54 NITROGEN (110)o + (220)an parent spots and in (b) the (002)oi. in the twin. The a" lamellae in R. D. Garwood^ and Gareth Thomas the parent crystal are less distinct in (a) because of their smaller inclination to the The tempering of iron-1.5% nitrogen foil plane (650 kV). The diffraction pattern martensice has been studied at temperatures (schematic) is shown in (c) while (d) is a up to 300°C, using X-ray and electron micro stereographic analysis of the twin and a" scope techniques. Stage 1 decomposition, lamellae traces (650 kV). (XBB 722-1254) occurs below 270°C by the general precipita tion, resembling spinodal morphology, of fii.s a" (Fei6N2) lamellae on {001} habit planes in boiii matrix and twin crystals of plates. It is believed that transformation the partially {112} twinned martensite plates occurs in this manner because the nucleation (Fig. 1). Yet, gauged by changes in the of coherent a" plates is controlled by the X-ray spectrum, the reaction is discontinuous, prevailing internal stress field. Thus the the tetragonal martensite doublets decaying time exponent "n" for the reaction decays in intensity without change in their Bragg from a normal value between 1 and 0.67 to positions (Fig. 2). The anomaly and the less than 0.3 as stress relief by recovery failure to detect by electron microscopy dominates the more protracted stages of the regions exhibiting fractional stages of the reaction. fine scale a1 -> a + a" reaction is attributed to its occurrence at different times in different martensite (or parts of martensite) Above 200°C the more staple nitride y' (Fe^) forms at an increasing rate as plates -112- 1 T" 30 min a'(IOI) (Oil) |lT»30sec T°°/\ I T°30seo/ f~\ 1 nl a"«(202> / / \\ URT=O (022) / / VI T = IOO0min/\ / / J8. \\ / \ / / Smin ft til LpJJ^\) a (1101 Y|\ *-T = 30 i i i i i"» 6 8 10 d-*A (mm ^l Fig. 1. Plot of stress intensity to initiate Fig. 2. Changes in the X-ray spectrum in the crack propagation and yield stress vs. (grain region of the (101)-(0U)a> martensite dou diameter) "V2, showing the linear dependence blets on tempering an iron-1.5% nitrogen al of these two parameters on loy at 150°C for the times indicated. d-1/2. Cobalt Kj radiation. (XBL 727-6600) (XBL 717-7014) This relationship is explained in terms on {012} habit planes, accompanied by marked of the microstructural differences between softening. the grain boundary regions and the matrix which results in the development of a triaxial stress state at the boundaries during deformation. The problem is analagous to LBL-470, Met. Trans, in press that with brazed joints. Although Cu-.Ni-Fe Department of Metallurgy, University College, alloys are relatively ductile, failure occurs Cardiff, Great Britain. preferentially at grain boundaries, because of the softer material there. 4. INTERGRANULAR FRACTURE OF AGED Cu-Ni-Fe Electron Microscopy and Structure of ALLOYS* Materials, edited by G. Thomas (U.C. Press, 19727: Ronald J. Livak''" and William W. Gerberich* +Part of Ph.D. thesis (LBL-1107) of R. J. Livak, now at Bell Telephone Research Labs. Cu-Ni-Fe alloys are subject to inter- N.J. granular cracking as a result of preferential coarsening of particles near the grain Now at University of Minnesota. boundary and also because of the discontinuous precipitation reaction which occurs. Figure 1 shows the linear dependence of the observed yield stresses and stress intensities on 5. SHORT RANGE ORDER IN Ni-Mo, Au-Cr, Au-V, grain size, which obeys the relation: AND Au-Mn ALLOYS* m Santosh K. Das,+ P. R. Okamoto,+ K e*>£ hucl. w> P. M. J. Fisher* and Gareth Thomas where fy is a constant, d is the grain diameter, The nature of short range order (sro) has e* is fracture strain at the grain boundary. been investigated in the systems Au4Cr, AujCr, Jl is the width of the grain boundary region, Ni4Mo, Ni3Mo, Au3Mn, and A114V by transmission E is Young's modulus. electron microscopy and diffraction. The -113- examination of the sro state, at temperatures above the critical temperature (Tc) for order-disorder transformation and also in the samples that were quenched from above Tc, showed diffuse scattering peaks centered near {1 1/2 0} positions in the fee reciprocal lattice, whose shape changed from one system to another. Evidence for diffuse scattering near superlattice positions corresponding to the Dla structure was obtained in the sro state of Au-Cr, Au-V, and Ni-Mo alloys. O FUNDAMENTAL FCC SPOTS Additional diffuse scattering was observed • DIO SCPEFB-ATTICE SPOTS O PROJECTION OF THE OTHER TJIO SPOTS near Ni2Mo positions in the Ni-Mo alloys. » R.,Vo SUPERLATTICE SPOTS Figure 1 shows schematically the diffraction •> PROJECTION OP THE li^O) SPOTS patterns that are observed. The diffraction patterns of Au3Mn above Tc showed evidence Fig. 1. (a) [001] reciprocal lattice section for diffuse scattering near the superlattice containing Dla, Iti^Mo, {1 1/2 0} sro spots; positions corresponding to Watanabe type (b) [130] reciprocal lattice section. two-dimensional long period superstructure (XBL 717-7009) (e.g., Fig. 2). Using the pairwise interaction parameters up to third nearest neighbors in the existing statistical thermodynamic model of Clapp and Moss, one can partially account for the shapes of {1 1/2 0} sro spots in Ni-Mo and Au-Mh alloys, but not in Au-Cr and Au-V alloys. On the other hand the diffuse sro scattering in all these alloys can be qualitatively explained by assuming the presence of imperfectly ordered microdomains or various superstructures that may exist in the long range ordered (tro) state in these alloys. For a particular stoichiometry —e structures Fig. 2. (a) [001] diffraction pattern of of the microdomains that exist above Tc may evaporated thin film of Au3Mn showing two- or may not correspond to the equilibrium dimensional long period superstructure S-ro structure obtained at that stoichiometric (Watanabe type), (b) another [001] pattern composition. showing additional Dla superlattice spots, (c) sketch of [001] pattern corresponding to Watanabe-type structure. (XBB 717-3414A) LBL-441, Acta Met., in press. Now at Argonne National Laboratory, Argonne, Illinois. Now at Melbourne, Australia. -114- D. HIGH VOLTAGE ELECTRON MICROSCOPY Gareth Thomas, Principal Investigator PENETRATION Introduction. High voltage electron microscopy has advantages in improved penetration and resolution and can be used to study directly knock-on damage in inorganic crystals. With increasing energy there is a limit to the useful thickness which can be penetrated if high resolution information is to be obtained. This appears to be ~ 1 MeV A 1 SiLiCON g-tll S t«mctic for most materials except very light elements. rf On the other hand, increasing the energy of r a SILICON g=izu " the electron beam reduces damage from inelastic • Q(,,STA(Nl£SSSTEEtg=l11 " and elastic scattering processes (e.g., ioniza tion damage). It appears therefore that very high energies may be beneficial for research on beam-sensitive materials (polymers, biological bodies). At the present time there are only two microscopes in the world (Toulouse, France and Osaka, Japan) which have capabilities up to 3 MeV. However, in mo.THOMAS. PHJL.HXG.!Z,10V|19I0) collaboration with Prof. R. M. Glaeser J.0 2.5 3.0 MCV (Medical Physics) we intend to make an in- depth study of the potential of HVEM up to, say, 5 MeV. It should be noted that it Fig. 1. Penetration data obtained for silicon may not be possible to produce stable beams and stainless steel. Open circles correspond above 3 MeV (and Wo parts in 10& volts is to thicknesses for which up to six fringes are necessary for instrumental resolution ~ 5A visible at stacking faults. Closed triangles or better), so that any gain in specimen correspond to total loss of fringes. The lifetime from reduced specimen damage may open triangle at 2.5 MeV corresponds to the be offset by poorer resolution. The costs thickness for which all fringes are just for a 3-5 MeV microscope would be very very visible. (XBL 728-6771) high. In inorganic materials an up-to-date The data are shown in Fig. 1, from which it survey of the current status of the theory is seen that for silicon the penetration and applications of HVEM is given in the increases up to about 14u, whereas for steel Proceedings of our 1971 IMRD Conference, little or no gain is apparent. Primary Electron Microscopy and Structure of knock-on damage occurred readily in both Materials, published by the University of materials. California Press, 1972. The critical voltages for the 440 Following are examples of current work in reflection in silicon and 422 in tantalum which emphasis is being placed upon improved were measured. Both were found to be resolution and penetration, enabling us to 1.4 MeV(± 10 kV). develop more structure-property programs on non-metallic systems. Measurements of radiation damage in valine and glycine by observing the disappearance of single-crystal spot diffraction patterns as a function of beam current and voltage 1. TRANSMISSION ELECTRON MICROSCOPY AT showed that the critical exposures for com 2.5 MeV* plete fading of the patterns increased with increasing voltages. Due to difficulties Gareth Thomac .and J. C. Lacaze+ in measurements of the current and end-point determinations, the magnitude of the increase Measurements of electron penetration on is not yet known with much accuracy. Work silicon and austenite stainless steels have is continuing on this problem, and at been continued, using the same criteria as Berkeley we plan to study lifetimes of these established previously [G. Thomas, Phil. Mag. materials by exposure to 5-15 MeV beams in XL, 1097 (1968)]. The work was done up to Linacs, followed by examination in our 650 kV 2.5 MeV, using the 3 MeV Toulouse microscope. microscope. -115- LBL-1119(Rev); J. Microscopy, in press; G. T. acknowledges support from a Guggenheim Fellowship. Lab. d'Opt.ique Electronique, Toulouse, France. 2. DEFECT CONTRAST AND RESOLUTION IN HIGH VOLTAGE ELECTRON MICROSCOPY M. J. Goringe,''' E. A. Hewat,^ C. J. Humphreys''' and Gareth Thomas Fig. 2. Dissociated edge dislocation profiles. The importance of several non-conventional (14 1/2, 14 1/2, 0) satisfied. 000, 220, and tecnniques in transmission electron 220 (dotted) images. Case for Cu. 1000 kV, microscopy for the improved imaging of t = 1634& at depth 817A, g.b = 1 for both defects has already been recognized. Signi partials. 's (XBL 731-5648) ficant benefits of improved imaging of dis location lines and quantitative applications to problems of disloction structure and stacking fault energy measurements have been At 100 W, dislocation images may normally demonstrated at 100 kV by the dark-field be understood by using two-beam theory which weak-beam technique (e.g., Cockayne et al., considers the two most strongly excited 1969). Comparably narrow dislocation images dispersion surface branches. As is well have been produced by high voltage electron known, the image width is then about ?g/3 microscopy, using bright-field images when where £~ is the extinction distance the crystal is oriented near high order [Cg = (2k)"1, where Ak is the separation of systematic reflections.1"3 Reasonably the two branches]. At higher voltages many- narrow, high contrast images have been pro beam interaction become more important, and duced in the second-order dark-field image for an understanding of defect contrast a when the_accelerating voltage is the critical larger number of dispersion surface branches voltage.3>4 The present program of research must normally be considered. Figure 1 is aimed at evaluating the advantages and illustrates the situation for a crystal disadvantages of these various techniques and oriented near the third-order Bragg reflecting to determine their fields of application for position. At A the third order is exactly the quantitative characterisation of defects satisfied. For the bright-field image, at in solids. 100 kV the transition 3 - 4 is the only one Reflection satisfied Fig. 1. Schematic dispersion surfaces illustrating important interband transitions caused by dislocation. A is for 3g, s = 0.B is 3g, s > 0. (XBL 731-5647) -116- dark-field image, such as used for weak-beam work, a single transition (2 - 4) is dominant. Similar reasoning applies to any order re flection along the systematic row, the 4k value of the dominant transition increasing with increasing order. The implications for microscopy are that both high-order bright- field and weak-beam images may be narrow-- see Fig. 2 (because of the large Ak values, and hence small ?g value, involved), but the bright-field image may be more complicated for quantitative analysis (two interactions Fig. 3. Deformed Si showing dissociated and rather than one), unless very high order decorated dislocation imaged 100 kV B.F reflections are used. (0,4g) s > 0 (g = 220). (XBB 731-265) Micrographs and computed profiles for specific cases are shown in Figs. 2-4. Points to note are the narrowing of both bright- and dark-field images with increasing order. In the bright-field case the limit to the order is set by lack of contrast. Very small image width is shown by the dis sociated dislocation of Fig. 3 (Ullage peak separation 90A). Note also the decoration effects after electron irradiation in the beam. The image profiles of Fig. 4 illustrate the improvements of the image at the critical voltage, both in terms of simpler image and higher contrast due to the low background. LBL-S41; Proc. EMCON 1972, Inst, of Physics (London), 1972, p. 538. Financial support is acknowledged from SRC and the Guggenheim Foundation (G.T.) Department of Metallurgy, University of Oxford, Oxford, England. 1. R. Osieckiand G. Thomas, in Proc. of 1971 EMSA Conf. (Claitors, Baton Rouge, 1971), p. 178. Fig. 4. Cu, edge dislocation. 222 satisfied, 2. F. M. C. Besag, I. P. Jones, and b -•* 1/2[110] (a) 470 kV; (b) 386 kV = Vc; R. E. Smallman, Jernkont. Ann. 155, 464 (1971). Tc) 300 kV. Thickness = 5.5?g, dislocation 3. W. L. Bell and G. Thomas, in~Electron depth = 3£g. [Notation: (g,ng) means image Microscopy and Structure of Materials (U. C. formed in Beam g with ng satisfied.] Press, Berkeley, 1972), p. 23. CXBL 731-5649) 4. J. S. Iaily, C. J. Humphreys, A.J.F. Metherell, and R. M. Fisher, Phil. Mag. 25, 321 (1972). of importance, but at 1 MV all the transitions indicated are important, because more Bloch waves are appreciably excited and more inter- band transition probabilities are large. 3. HIGH VOLTAGE ELECTRON MICROSCOPIC STUDIES Since different interactions can give rise OF MINERALS* to images in different positions, a compli cated image may res'ilt. Note however that Wolfgang F. Muller, H.-R. Wenk,+ and at the critical voltage the image is simpli Gareth Thomas fied. In particular, image tails due to 2-3 transitions will be absent. At B a. On the Transformation Qrtho-enstatite to (positive of third-order Bragg position) the CI ino -enstatite situation is less complicated but still involves at best two interactions for the The ortho-clino inversion in enstatite bright field image. For the first-order (MgSi03) is interesting because the trans- .1. Wenk (1968) and favored in the mechanism of Coe (1970). It contradicts the transforma tion mechanisms suggested by Turner et al. (1960), Brown et al. (1961) and Starkey (1968). Joint program with Geology. Fig. 1. High voltage electron microscopy of Associate Profc=-.jr, Geology. undeformed and deformed Bamble enstatite. Left: BF electron micrograph of an unde formed specimen showing ortho-enstatite with b. Order in Anorthite* stacking faults parallel to (100]. Center: BF electron micrograph of the deformed speci Calcic plagioclase feldspars with a chemi men showing alternating lamellae parallel to cal composition ranging from (100) of untwinned clino-enstatite and ortho- Cao.9Nao.lAl1.9Si2.lO8 (An 90) to CaAl2Si208 enstatite. Right: Selected area electron (An 100) aTe called anorthites. Two different diffraction pattern corresponding to the types of APBs may occur: Type b APBs (imaged micrograph in center showing the relation of with b-reflections: h + k odd, 1 odd; first the ortho- and clino-enstatite reciprocal observed by Christie et al., 1970) and type axes to each other and to the compression c APBs (imaged with c-reflections: h + k directions (heavy arrows). 6S0 kV accelera even, 1 odd). Presence and size of antiphase tion voltage. (MB 727-3691) domains depend on the chemical composition and the cooling history of the anorthite. A comparative study of terrestrial, roeteoritic, and lunar anorthites with the formation from the orthorhombic to the mono- high voltage microscope showed that the size clinic polymorph is dramatically promoted of domains increases with increasing An- by shearing on (100) planes parallel to [001], contents in the slowly cooled metamorphic but is only slightly promoted by hydrostatic and plutonic rocks. Type c domains were small compression. Earlier investigators found or not resolvable in volcanic, i.e., quenched, that clino-enstatite (CE) formed during rocks (compare electron micrographs f and g deformation has the |a| and |8| crystal- in Fig. 2; the chemical composition is the lographic axes in common with the parent same in the volcanic and meteoritic ortho-enstatite (OE) and has (100) as habit anorthite). plane. But they could not deduce the position of the jc| axis uniquely. The electron Type b-APBs were absent in all terrestrial microscope study was done on a sample from anorthites studied so far which had chemical a Bamble-enstatite specimen shortened 3.7? compositions ranging from An 94 to An 100. at a strain rate of 10"4 sec"l with a con They were present, however, in the calcic fining pressure of 5 kbar and a temperature plagioclase (about An 85 to An 95) from of 800°C. This sample was provided by lunar basalt 14310 (Fig. 2, first line). Prof. R. Coe, U.C. Santa Cruz. The prepara Most of the b-APBs were 500 to 1000 A wide, tion method, i.e., ion-thinning of a pre but in rare cases larger b-domains have been selected area, enabled correlation to be observed in the same specimen. No c-domains made of the electron micrographs and dif have been resolved in anorthites from 14310. fraction patterns to the oriented petrographic thin section, so that the compression direc The antiphase vector of c-APBs was tion of the deformation experiment is known analyzed as 1/2 [&_ + b + c]. Figure 3 shows on the micrographs. Electron microscopy c-APBs in anorthite from val Pasmeda, S. Tyrol revealed submicroscopic lamellae which are (An 100). The fringes have the properties alternately OE and untwinned CE in contact expected for the phase factor a = it, thereby on (100) habit planes and with axes a_ and 6 providing additional proof for the antiphase parallel (|b0El = IbCEl, |aOEl = 2 |acEl). vector 1/2 [£ + b_ + cj. Calculated intensity Figure 1 displays electron micrographs of profiles applying multi-beam dynamical theory undeformed enstatite [containing stacking are in agreement with the observed fringe faults parallel to (100) in some areas] and patterns. of the deformed enstatite; the orientation of the £ axis of CE with respect to the com pression direction is shown in the diffraction a pattern. It is the orientation made likely Mineralogy and Petrology 34, 304-314 (1972). by the optical measurements of Trommsdorff and (LBL-460). — -118- FHflOMICtOCIUMt, IPOL X -IH. PIECESSIO* ELCCrroHIIFFUCTIOM ELECTMKMICIOCMPH Fig. 2. Structural variations in anorthites. Is* line: lunar basalt 14310; 2nd line: volcanic tuff, Japan; 3rd line: meteorite Serra de Mage; 4™ line: metamorphic calesilicate rock, Alps. The electron micrographs are dark field images taken at 650 kV acceleration voltage. (XBB 7112-6108) Fig. 3. Type c APBs in anorthite from S. Tyrol in bright and dark field. 650 kV acceleration voltage. (XBB 728-4151) -119- 4. HIGH ANGLE TILTING STAGE FOR THE 650 kV trast experiments. Stacking faults on {110} HITACHI MICROSCOPE* have been found in the cation sublattice and also (and unexpectedly) {111} faults on the Marcel Bouchard, William L. Worthington, oxygen sublattice. Donald J. Jurica, and P. R. Swann'f' Dislocations in lithium ferrite are A new tilting stage has been developed dissociated into two partials of Eurgers which allows ±30° tilt in any direction. vectors a/4 (110 > . These define a stacking This stage does not require any modifications fault in the cation sublattice. to the microscope. Figure 1 shows the dia gram of the new stage. Present work is concentrated on a more quantitative analysis of the defect structures and precipitation reactions including studies of the interphase interfaces. LBL-1145; Rev. Sci. Instr., in press. Reader in Metallurgy, Imperial College, London, England. 6. OTHER RESEARCH PROJECTS IN PROGRESS -TEFLON RINO F Gareth Thomas a) Phase Transformations (i) Bainitic transformations in a range of steels have been studied. The results indicate that there is no sharp transition between upper and lower bainite. Epsilon carbide has been identified in lower bainite; bainite can also contain transformation twins, i.e., bainites can have the same range of structural features as tempered martensites. (ii) Structure and mechanical properties have been studied in Cu-Mn-Al alloys some of which undergo order + spinodal reactions. These alloys are useful in magnetic applica tions but are mechanically brittle. (iii) Transformations in ceramics and minerals. Studies of order-disorder in lithium ferrite show existence of LRO with a/2 <110 >a.p.b.s. In addition, faulting has been resolved both in the anion and cation Fig. 1. a) Schematic cross section of the sublattices. Order has also been investigated tilting stage. in anorthites and two different types of b) Magnified portion of the tilting a.p.b. have been detected in both terrestrial mechanism showing the sphere in the zero (0°) and lunar samples. tilt position. c) Magnified portion of the tilting b) Contrast and Resolution in HVB1 mechanism showing the sphere in the maximum Unconventional techniques such as high (30°) tilt position. (XBL 728-6878) order bright field imaging or weak beam dark field imaging in systematic reflections have been shown to improve lattice defect resolu tion considerably. Work is being done to 5. LITHIUM FERRITE resolve closely spaced dislocation pairs and, in a cooperative program with Professor J. Omer Van der Biest Washburn, to interpret contrast from defects in ion-implanted silicon. Work is progressing on characterizing the structure of LiFesOs in which studies are c) Martensite Formation in Single Crystals being made of the order-disorder transition. Anti-phase domain boundaries with Mechanism of strain-induced transforma j5 = a/2 < 110 ) have been identified by con tions and mechanical properties of metastable -120- austenitic stainless steels are being 2. I-Lin Cheng and G. Thomas, The Enhance investigated in single crystals of various ment of Strengthening Dislocated Marter-site, orientations so as to obtain single and Met. Trans. 3_, 503-516, 1972. (UCRL-20562). multiple slip systems in austenite. 3. W. L. Bell and G. Thomas, Applications and Recent Developments in Transmission Electron Microscopy, in Electron Microscopy 7. RESEARCH PLANS FOR CALENDAR YEAR 1973 and Structure of Materials, (Univ. of Calif. Press, Berkeley) pp. 23-59. Gareth Thomas 4. W. F. Muller, H.-R. Wenk and G. Thomas, a. Alloy Strengthening Structural Variations in Anorthites, Mineralogy and Petrology 3£, 304-314, 1972 Continuation of program on Fe-Cr-C and (LBL-460). Fe-Ma-C steels with emphasis on improving strength and toughness. Program will include 5. R. Osiecki and G. Thomas, High Voltage studies of lower bainites corresponding to Electron Microscopy of Interfac al Defects those alloys which are brittle in the mar- in GaAs, Materials Sci. and Engr. 10 (1), 53, tensitic state. 1972 (LBL-440). Work will be expanded to cover a wider 6. M. J. Goringe, E. A. He-'at, C. J. range of alloys in the Cu-Mn-Al system so as Humphreys and G. Thomas, Defect Contrast and to better understand the phase relationships Resolution in High Voltage Electron and properties. Further studies on spinodal Microscopy, Fifth European Electron Fe-Ni-Cu will be done in an effort to prevent Microscopy Conference 1972, (Institute of the grain boundary reaction and embrittlement. Physics London 1972) p. 538 (LBL-841). b. Non-metals 7. M. Bouchard, R. J. Livak and G. Thomas, Interphase Interfaces in Spinodal Alloys, Work on mineralogy will be slower due to Surface Science 31, 275-295, 1972 (LBL-152). the departure of W. F. Muller, but we hope for extra funding from NSF to keep the joint 8. G. Thomas and C. Vercaemer, Enhanced program with Professor Wenk (Geology) going. Strengthening of a Spinodal Fe-Ni-Cu Alloy by Martensitic Transformation, Met. Trans. Current work on spinels will be continued. 3, 2501-2506, 1972 (LBL-447). Searches will be done for spinodal systems. Studies of plastic deformation and phase 9. D. Huang and G. Thomas, Discussion: transformations in minerals (joint program Structure and Mechanical Properties of Tem with Prof. Wenk, Geology) will be continued. pered Martensite and Lower Dainite in Fe-Ni-Mn-C Steels, Met. Trans. 3 (1), 343- c. New Projects 346, 1972. New projects under way on a joint program 10. Electron Microscopy and Structure of with Prof. R. M. Glaeser (Medical Physics) Materials. Proceedings of the >i±th Intem- include studies of high energy electron ational Materials Symposium "The Structure irradiation damage in organic/biological and Properties of Materials--Techniques and materials, e.g., Linac work on amino acids Applications of Electron Microscopy," held (5-15 MeV), viruses, etc., and evaluation at the University of California, Berkeley, of the potential of 5-10 MeV electron Sept. 13-17, 1971 edited by Gareth Thomas microscopy. (University of California Press, Berkeley, Los Angeles, London) 1972. 11. Masatoshi Fukamachi, Investigations of Images of a Superdislocation and a Disloca 8. 1972 PUBLICATIONS AND REPORTS tion Dipole Using the Systematic Reflections, Japanese Journal of Applied Physics 11 (10), Gareth Thomas and Associates 139V1402, 1972. Journals and books 12. H. R.-Wenk, M. Ulbrich and W. F. Muller, Lunar Plagioclase: A Mineralogical Study, in 1. M. N. Chandrasekharaiah, S. Ranganathan, Proceedings of the Third Lunar Science Confer P. R. OkajiDto and G. Thomas, Field-Ion ence (Supplement 3, Geochimica et Cosmochimica Microscopic Study of Antiparallel Twins in Acta) Vol. 1, pp. 569-579, The M.I.T. Press, M.4M0, Mat. Res. Bull. 7, 13, 1972 (LBL-412). 1972. -121- LBL reports 7. Ronald John Livak, Spinodal Decomposition and Coarsening in Cu-N'i-Fe Alloys (Ph.D. 1. S. K. Das, F. R. Okamoto, P. M. J. Fisher thesis) Sept. 1972. and G. Thomas, Short Range Order in Ni-Mc, Au-Cr, Au-V and Au-Mn •Alloys. LBL-441 Rev. 8. John McMahon^ Structure, Strength, and December 1972, Acta Met., in press. Toughness of Fe/Cr/C Martensitic Steels, (M.S. thesis) LBL-1181, Dec. 1972. 2. R. D. Garwood and G. Thomas, The Tempering of Martensitein an Iron 1.5" Nitrogen Alloy, 9. K. Seshan, W. L. Bell and J. Washburn, (Met. Trans, in press) LBL-470, Aug. 1972. A Rapid Method for Distinguishing Vacancy and Interstitial Loops in Ion Implanted 3. G. Thomas and J.-C. Lacaze, Transmission Crystals, submitted Physica Status Solidi, Electron Microscopy at 2.5 MeV, accepted J. LBL-1103, August 1972. of Microscopy, LBL-1119 Aug. 1972. 10. M. Bouchard, h. L. Worthington, D. J. 4. P. R. Swann, G. Thomas and N. lighe, Jurica and P. R. Swann, A High Angle Tilting In-Situ Observations of the Nitriding of Stage for the 650 kV Hitachi Electron Tantalum, ProceeCngS of Symposium on High Microscope, (accepted by Rev. of Sci. Instr.) Voltage—Electron Microscope held in LBL-1145, Oct. 1972. Manchester, England Sept. 1972. (Royal Microscopical Society--in press). 11. W. F. Muller, H. R. Wen);, IV. L. Bell and G, Thomas, Analysis of the Displacement 5. Teh Y. Tan, Behavior of Inelastic Vectors of Antiphase Domain 'lundaries in Electrons, (Ph.D. thesis) LBL-432, January Anorthites (CaAl2Si208), 1G72. 6. Mathur V. Raghavan, Structure and Mechani 12. M. von Heimendahl, Specimen Sickness cal Properties of Fe-Ni-Co Alloys with and Determination in Transmission Electron . ithout Carbon, (D. Eng. thesis) LBL-*77, Microscopy in the General Case, LBL-I186, March 1972. Oct. 1972, Micron in press. -122- E. ALLOY DSSIGN Victor P. Zaakc.y and tarl R. Parkbr-, 1. THE FRACTURE TOUGHNESS OF ULTRA-HIGH Principal Imsestigavora STRENGTH ALLOY STEELS INTRODUCTION" William E. Ifcod The application of the first principles This investigation showed that the tough- oi" materials science Co dsvelop nc; and cess of several commercial low alloy high improved alloys is a relatively recent strength steels is a sensitive function of approach to the problem of alloy design. their micrcstructure snd heat treatment. It Virtually all of the presently available was also concluded that the commonly used caimercial alloys have been discovered by Charpy-V-notch test is not sensitive to small trial-and-error methods. This approach is changes in microstructure which can drastical no ionger feasible because c?f its inefficiency ly alter fracture toughness. and expense. Further, such an approach is totally inadequate to inset the complexity Tits steels ured in this study were AISI of material requirements of present and 4130, 4330, 4340, 4140, and 300-M. Their future technology. compositions are listed in Table I. The normal commercial heat treatment for all the Mechanical properties of alloys largely above steels consists of austenitizing for depend on chemical composition and structural 1 hr at 870°C followed by oil quenching. features such as crystal structure, defect However, austenitizing at temperatures up to structure, and microstructure. The develop 1200°C (and for some steels by quenching into ment of a unified theory of fracture tough solutions as severe as ice brine) followed by ness has done much to relate quantitatively refrigeration in liquid nitrogen, a drastic the variables of chemical composition and diange was observed in the fracture toughness. microstructure to fracture toughness. The high resolution instruments needed to analyze Figure 1 shows the room temperature plane composition on a microstructural scale and strain fracture toughness of AISI 4130 as to characterize; elements of structure whose a function of tempering temperature for both dimensions are of the order of a micron or the commercial heat treatment and the 1200°C o. fraction of a micron are now readily austenitizing treatment followed by an ice available. It is our belief that the brine quench and liquid nitrogen refrigeration availability of these instruments coupled with (IBQLN). An increase in fracture toughness an understanding of the principles of of 100% was observed for the latter treatment dislocation theory, thermodynamics, alloy over the former. This increase was associated chemistry, and micromechanics of deformation with a microstructural change from a mixed and fracture will aid in the development of microstructure of blocky fcrrite and superior alloys. Examples of present efforts directed toward this goal include the following: 1) Study of these chemical and micro- structural features x.'hich control the mechanical properties if low alloy ultra-high strength steels. 2) The design, froiii first principles, of an interstitial-free cryogenic alloy. 3) The design and development of a non- carbon-containing precipitation hardening ferritic alloy for elevated temperature applications. 4) The design of new TRIP steel processes which involve a minimum of prior austenite deformation. 5) The design of high manganese-high carbon steels utilizing a strain induced transformation. Concurrent with the studies mentioned Fig. 1. The effect of tempering temperature above has been the development of a rapid on the room temperature plane strain fracture magnetometric technique for detecting the toughness, kjrj, of AISI 4130, austenitized early stages of isothermal transformations and quenched as indicated. (XBL 731-5651) in steels. -123- Table I. Chemical compositions of steels. Steel type Mn CT Ni Mo Si AISI 4130 0.30 0.46 0.85 0.15 0.20 0.011 0.28 AISI 4330 0.30 1.02 0.85 1.80 .40 .007 .28 0.05 AISI 4340 0.41 0.80 0.72 1.70 .24 .011 .23 300-M 0.40 0.79 0.75 1.85 .43 .003 l.SP 0.09 using the same heat treatment, Fig. 5. How ever, most notable was the observation that 4330 did not exhibit temper embrittlement at tempering temperature above 200°C. The absence of temper embrittlement was believed to be due to the presence of a small amount of vanadium, which prevents the formation of carbide films. Scanning electron fractographs of embrittled 4130 tempered for 1 hr at 3S0*C after an ice brine quench from 1200°C are shown in Fig. 6. The scanning electron fractograph for non-embrittled 4330 with the same heat treatment is shown in Fig. 7. arc Fig. 2. Microstructure of AISI 4130, oil quenched from the austenitizing temperature of 870°C, showing blocky ferrite and martensite. (XBB 731-377] martensite, Fig. 2, to one that was completely martensitic, Fig. 3. As seen from Fig. 1, a heat treatment of 1200°C followed by an oil quench resulted in an intermediate level of fracture toughness. The microstructure associated with this heat treatment was mostly martensite but with regions of upper bainite along prior austenite grain boundaries. (See Fig. 4). Fig. 3. Microstructure of AISI 4130, ice brine quenched from the austenitizing tem The steel AISI 4330 was similar to 4130 perature of 120CC. The structure is almost in that the best results were attained by completely martensitic. (XBB 731-374) Fig. 6. Scanning electron fractograph of Fig. 4. Microstructure of AISI 4130, oil AISI 4130, austenitized at 1200°C for 1 hr, quenched from the austenitizing temperature ice brine quenched, refrigerated in liquid of 1200°C, showing upper bainite at the prior nitrogen, and tempered at 3S0°C for 1 hr, austenite grain boundaries, and martensite. and tested for kjc at room temperature. (XBB 731-375) Fracture occurred by intergranular cleavage. (XBB 731-375) Whereas the 4330 had a very ductile fracture 1 1 . 1 1 1 1 mode, the 4130 exhibited an intergranular 0 cleavage mode of failure. ~ ——-—~~ r ' ~ The higher carbon steels also showed a -"^SoO "C, ICE BRINE QUENCH marked improvement in toughness although the - - optimum properties could be achieved with an oil quench. The 300-M steel was similar to 4330 in that it did not embrittle. The plane E strain fracture toughness increased when the tempering temperature was raised, as shown in - - Fig. 8. Finally, steel 4340 which did not i- •^970 "C, OIL QUENCH have any vanadium did exhibit an embrittle - ment on tempering at temperatures above 300°C, • V as indicated by the variation of plane strain fracture toughness with tempering temperature - (Fig. 9). In the case of 4340 a straight t 1 1 l quench from 1200°C caused quench cracks; therefore, the steel was quenched first to 870°C an>_ then into oil. A result that was Fig. S. The effect of tempering temperature common to all the steels was the increase in on the room temperature plane strain fracture grain size from ASTM 7-8 to 0-1 when the toughness, iqc, of AISI 4330, austenitized austenitizing temperature was raised from and quenched as indicated. (XBL 731-5652) 870 to 1200 °C. -125- Fig. 9. The effect of tempering temperature on the room temperature plane strain fracture toughness, Iqc, of AISI 4340, austenitized and quenched as indicated. (XBL 731-5654) Finally impurity elements arid inclusions are well-known sources of poor toughness. Investigations are in progress to determine the role of both the amount and distribution Fig. 7. Scanning electron fractograph of of impurity elements and inclusions on the AISI 4330, austenitized at 1200°C for 1 hr, fracture toughness of AISI 4340. ice brine quenched, refrigerated in liquid nitrogen, and tempered at 350°C for 1 hr, and tested for kic at room temperature. The mode of failure was dimpled rupture. 2. STRUCTURE AND FRACTURE TOUGHNESS OF (XBB 731-373] SECONDARY HARDENING STEELS Thomas Tom The relation between microstructure and fracture toughness was studied in several secondary hardening steels of varying carbon, molybdenum, and nickel contents. Ihe austenitizing temperature, the tempering temperature, and the chemical composition were found to have a considerable influence on the plane strain fracture toughness, KIQ, by virtue of their role in changing microstructure. Fig. 8. The effect of tempering temperature on the room temperature plane strain fracture jl !—i 1 i I i I i I toughness, kic, of the 300-M steel, O AS 100 POO 300 400 a£NQ D austenitized and quenched as indicated. * TEMPERING TEMPERATURE ft) CXBL 731-5653) -126- The nominal compositions of the steels Table I. Chemical compositions of steels. are listed in Table I. The heat treatment consisted of an hour austenitizing treatment at several temperatures from 1600 to 2200°F Compositions, wtS followed by ice brine quenching and refrigera tion in liquid nitrogen to insure complete Designations C Mo Ni Mn transformation to martensite. Que hour tempering treatments at 400 and 575°F were A 0.3 5 0.6 investigated for the 2200°F austenitizing treatment. B 0.4 5 0.6 The room temperature plain strain fracture C 0.3 4 3 0.6 toughnesses of steels A and C are plotted in Fig. 1 as a function of the austenitizing D 0.4 4 3 0.6 temperature. The KJC value increased considerably when the austenitizing tempera ture was raised to 2000°F and above. The higher Kic for the higher austenitizing temperatures were attributed to the elimination of undissolved carbides, a more favorable distribution of impurities such as sulphur and phosphorous, and to auto- tempering (depletion of carbon from solution). 120 z 100- ;*: ,i 1600 BOO SOOO 2200 AUSTENITIZING TEMPERATURE r>) Fig. 1. Influence of austenitizing tempera Fig. 2. Fractographs of steel A austenitized ture on the room temperature plane strain at (a) 1600°F and (b) 2200°F, ice brine fracture toughness of secondary hardening quenched and tested at room temperature for steels A and C tested in the as-quenched plane strain fracture toughness, condition. (XBL 731-5623) (XBB 731-372) -127- 100 i 1 1 Fracture featcies of quasi-cleavage were ._,. observed for the 2200°F austenitizing treatment, Fig. 2(b). An increase of room temperature Kic with austenitizing temperature was also observed for steel B (See Table I below), as shown 80 -- in Fig. 3. Also plotted in the figure are 0.4 C, 5 Mo, 0.6 Mn-, the Kjc values for steel D which showed little change with increase in austenitizing _ temperature. The reasons for the unexpected - behavior of steel D are being sought. Steels austenitized at 220O°F and ice 60-_ / brine quenched were tempered for one hour at both 400 and 575°F and the room tempera ture Kic values determined, see Table II. / w Table II. The effect of tempering on the room temperature plane strain fracture toughness (ksi-in.172) of steels austenitized at 2200°F and ice brine quenched. As Steel austenitized Tempering temperature (°F) designation and quenched 400 575 A 100 109 107 B 70 101 93 C 113.5 128.3 118 D 55.2 96.9 -128- Table I. Fatigue crack growth properties. A. Fe-•5MO-0.3C Steel --Austenicized at 2200°F AK (ksi-in.1/'2) for da/dN (in/cycle) Tempering Specimen Temperature, Co n 5 x 10"6 1 x 10"5 5 x 10"5 aC(°F) S-3-1 as quenched 1.87xl0"15 2.2 22.40 31.30 64.40 8-3-2 225(425) 1.31xl0"14 2.0 21.90 31.60 72.50 5-2-3 300(572) 1.85x10"15 2.1 24.00 33.90 70.80 8-2-4 500(932) 1.06xl0"15 2.3 15.90 21.90 43.60 B. hi-5Mo-0.3 Steel --Austenitized at 1600°F 1/2, AK (ksi-in. 'n for ia/dN (in/cycle) Tempering Specimen Temperature, 5 x io"6 °C(°F) 21 18-6-2 as quenched 1.47x10 3.5 32.40 39.80 63.10 •20 18-6-3 225(435) 9.16x10 3.1 30.20 38.00 63.90 -20 18-6-6 300(572) 1.97x10 3.2 30.90 38.90 63.90 15-6-8 500(932) 2.09x10" 3.2 29.90 37.20 62.50 -20 18-6-0 600(1112) 3.50x10 3.2 25.40 31.60 41.20 martensite structure and the subsequent attributed to a difference in the mechanism tempering reactions. for crack growth. Changes in growth rates were also observed with changes in tempering Preliminary observations of fatigue crack temperature; high tempering temperatures had propagation indicated distinct differences adverse effects. in crack growth rates for the two austenitiz- ing treatments. Crack growth rates, da/dN Research in progress is aimed at obtaining were related to stress intensities, AK, by more crack growth data in order to character the equation ize microstructural features with mechanisms of crack propagation and fracture. m - c° wn 1. R. D. Gcolsby, Relationships Between The coefficients Co and exponents n were Microstrueture and Fracture Toughness in a estimated. The data aiesummarized in Table I. Secondary Hardening Steel (Ph.D. thesis), The distinc* difference in the exponents for University of California, Berkeley, LBL-405, the two aus.3nitizing treatments was Nov. 1971. -129- 4. THE INFLUENCE OF COOLING RATE IN A curves have been based on the cooling rate JOMINY TEST ON TOUGHNESS OF STEELS at 1300°F, because the cooling rate at this point determines whether or not the cooling Roy N. Lott curve passes through the pearlite knee in the TTT diagram of the steel. However, The objective of this study was to examine austenite to bainite and/or to martensite critically the Jominy end quench test as a transformations occur below the pearlite design criterion. Jominy curves of steels knee and the cooling rate at lower tempera are used to predict properties of quenched tures should influence the final properties parts. The measured hardness at any point of tho steel by its effect on the amount and of interest on the part is transferred to the type of bainite and martensite formed. Jominy curve to obtain the cooling rate at 1300°F. From the hardness and the cooling In the present investigation, the end rate at 1300°F., attempts are made to predict quench specimen was modified to a rectangular the properties of the quenched part. Jominy bar so it could be cut into Charpy V-notch bars after the end quench heat treatment, Fig. 1. These involve cooling curve deter NOTCH minations for bar center and half-radius positions in bars of diameters 1 in., 2 in., 2 1/2 in., and 3 in. Cooling rate vs bar Z) diameter and cooling rate at 1300°F vs cooling cz rate at a temperature in the bainite region 8 can be determined from the above experiments. Following the same procedure for cooling curves obtained at points on the end quench specimen, a comparison of the curves should 7 help indicate the fallacy of prediction of microstructure and properties from the cooling 6 rfc 1 ! STANDARD CURVC 5 4 l/„" 4 » 2 ' ' QUENCHED ENDT r NOTCH COOLING RATES AT 700 F 0.394" 10 20 DISTANCE FROM QUENCHED END I'/lS INCH] 2.164"- Fig. 2. Cooling rate at 1300 and 700°F vs distance from quenched end in the modified Fig. 1. Sketch of the modified end quench Jominy bar of the present investigation. specimen showing the location of the Charpy Also shown are the 1300°F data for a standard V-notch bars cut from it. (XBL 731-5655) Jominy bar. (XBL 731-5656) -130- Table I. Charpy toughness data for AISI 4130 steel as a function of test temperature and position along the end quench bar (specimen numbers correspond to positions as indicated in Fig. 1). Specimen Charpy V-notch energy fft lb) Heat treatment No. at test temperature (°F). RT 150 212 392 1 10.6 23.7 17.5 20.5 2 16.9 55.1 71.5 93.2 3 17.9 78.5 95.0 Standard 4 17.7 67.0 77.0 quench 5 14.5 42.0 57.5 65.0 6 17.3 42.0 55.5 59.0 7 19.3 39.2 57.1 61.5 8 20.0 36.5 61.4 64.0 Specimen No. RT 150 212 592 1 10.0 15.0 12.0 14.0 2 5.1 11.0 17.5 32.5 3 6.6 17.0 21.0 50.0 Step 4 8.0 12.5 24.5 52.5 quench 5 8.0 11.2 31.5 55.5 6 6.9 19.5 29.0 54.5 7 7.7 13.2 30.0 54.5 8 6.2 16.0 31.5 55.0 1 1 1 60 4340 " w in "^^^^ HI a SO < \ \ STANDARD Fig. 3. Hardness vs distance from quenched end of the modified Jominy bar of AISI 4340 steel austenitized and quenched as indicated. STEP 40 QUENCH " The broken lines are the specified limits of hardness for a standard Jominy bar of this steel. (XBL 731-5657) I C> 1 2 3 4 DISTANCE FROM QUENCHED END (INCHES) -131- Table II. Qiarpy toughness data for AISI 4340 steel as a function of test temperature and position along the end quench bar (specimen numbers correspond to positions as indicated in Fig. 1). Specijnen Qiarpy V-notch energy (ft lb; Heat treatment No. at test temperature (°F) RT 150 212 392 1 4.2 5.2 5.0 16.0 2 6.3 6.2 7.5 10.6 3 6.2 5.7 7.0 9.4 Standard 4 5.9 5.4 7.0 9.0 quench 5 5.4 6.2 6.4 12.5 6 5.5 6.5 6.4 21.5 7 5.6 6.4 8.0 21.0 8 5.5 6.6 8.0 23.0 Specimen No. RT 150 212 392 1 5.0 6.8 8.0 13.5 2 6.9 7.8 9.S 11.5 3 6.0 7.2 8.5 10.5 Step 4 5.5 12.7 7.5 9.0 quench 5 6.9 8.8 10.0 12.5 6 6.6 10.3 13.0 21.0 7 9.1 7.4 11.0 27.5 8 5.8 7.4 11.0 30.0 i i 1 4130 - ~ \ " AN-— STEP — nneuru Fig. 4. Hardness vs distance from quenched STANDUS"" end of the modified Jominy bar of AISI 4130 QUENCH steel austenitized and quenched as indicated. The broken lines are the specified limits - " of hardness for a standard Jominy bar of this steel. (XBL 731-5SS8) 1 1 DISTANCE FROM QUENCHED END CINCHES) -132- rate at 1300°F. Specimens were quenched in two classes: 1) those that show very high an end quench rig modified for rectangular strength but low toughness and which specimens with a slit instead of a round generally have a body centered cubic matrix, nozzle. The behavior of two steels, AISI 4130 and 2) those that show high toughness but and 4340, was examined in the present in relatively low strength and have a face vestigation. centered cubic structure. Our initial approach has been to investigate bcc iron- It has been shown that the toughness of based alloys with the purpose of developing AISI 4130 and 4340 steels are greatly exceptional combinations of high strength improved by austenizing at 2200°F and step as well as high resistance to catastrophic quenching to the commonly used austenizing brittle failure at low temperatures. This temperature (1600, 1550°F respectively) research aiproach reduces to the fundamental before quenching them to room temperature. problem of decreasing the ductile-brittle Accordingly, one set of specimens of ead. transition temperature in basically body alloy was step quenched (1 hr at 2200°F •* centered cubic structures. A number of 1/2 hr 1S50 or 1600°F * end quench) and the factors are known to influence the low other set was given the standard heat treat temperature toughness of bcc iron and its ment (1/2 hr 1550 or 1600°F •* end quench). ductile-brittle transition temperature. Of The rectangular end quench specimens of both these the role of alloys conformed closely to the standard Jominy end quench as demonstrated by the 1) interstitial elements, cooling rate curves in Fig. 2 and the hard 2) substitutional elements, ness curves in Figs. 3 and 4. Charpy V-notch 3) grain structure, and bars cut from the end quenched specimens 4) dislocation substructure (see Fig. 1) were broken at temperatures selected so that the ductile-brittle transi were felt to be of primary importance. Based tion curve could be determined for each on the above considerations a Fe-12SNi-0.5%Ti position on the end quenched specimen for alloy was selected for study. each alloy and heat treatment. Table I shows the Charpy-V-notch values of AISI 4130 steel The addition of 128Ni (factor 2 above) in both the standard quenched and step quenched conditions. Table II shows the Charpy test data for AISI 4340 steel. Experiments to determine the cooling curves of round bar sections quenched in agitated oil and agitated water are now in progress. 1. W. E. Wood, Ph.D. thesis, University of California, Berkeley (LBL-1474). 5. DESIGN OF INTERSTITIAL-FREE CRYOGENIC ALLOYS Michael J. Yokota, Wade A. Horwood, and Go Sasaki The mechanical behavior of metals at cryogenic temperatures is of technological importance as well as of scientific interest. With the rapidly expanding use of liquified gases in scientific and industrial applica tions there is a great demand for materials which are both tough and strong at tempera tures down to -200°C. Examples include soo' 1— 1 container material for propellants used in O 10 SO 30 space and missile systems and for use in the WEI8HT PERCENT Nl M ft-SjTI storage and transportation of liquid natural Fig. 1. Portion of a tentative phase diagram gases. of the Fe-Ni-Ti system showing the trans formation temperatures for the Fe-12Ni-0.STi Presently available materials fall into alloy. (XEL 731-5660) -133- serves several functions. A unique toughening at low temperatures. A further improvement is produced in nickel containing ferritic in notch toughness could be attained by steels at low temperatures. Only manganese optimizing the microstructura features such and, possibly, several elements of the as grain size and dislocation substructure platinum group have comparable effects on (factors 3 and 4 above) through thermal and iron* Secondly, with the addition of 12$Ni, mechanical treatments. the equilibrium two phase a+y field widens abruptly while also being lowered by several The notch toughness of several commercial hundred degrees, Fig. 1. This feature of the bcc iron-based alloys are modestly improved alloy allows for possible grain refinement by a reheating treatment usually several (factor 3 above) by a phase transformation- hundred degrees below the equilibrium recrystallization process, which is part of austenite phase field. A number of reasons the present investigation and is discussed have been given to explain this phenomenon in a later section. but none seem to apply generally in all cases. Figure 2(a) shows the effect of reheating Titanium was added to act as a scavenger temperature on the energy absorbed at liquid for impurities, especially interstitials N2 temperature by standard Charpy V-notch (factor 1 above) and also to achieve specimens of the Fe-12*Ni-0.58Ti alloy. strengthening via solid solution and precipi Specimens were initially sustditized at tation hardening. A low free-interstitial 900°C for two hours ind air cooled. The content was felt to be of primary importance reheating temperature corresponding to the in producing good notch toughness at low maximum Charpy V-notch energy was in the temperatures. Additional strengthening by temperature range for rapid martensite to Ti was also felt to be desirable to compensate austenite transformation. for the removal of carbon in its usual role as a hardening element in steels. Optical examination of the 700 and 750°C reheated structures showed that grain refine The purpose of the above considerations ment from an austenite grain size of 50 um was to produce a matrix structure which would to a reheated grain size of 10 ym occurred exhibit good resistance to brittle fracture during the a -<• Y transformation. The grain GRAIN SIZE (/im) HARDNESS AT 2>°C (R03KWELL A) 10 20 30 40 50 56 58 60 G2 64 r; 900 -i—i—i—r 1 T CHARPY V-NOTCH UJ 'ENERGY ,' • "V GRAIN r- < 800 SIZEJ a: u a. 2 UJ 700 S 600 0 50 !00 150 200 130 140 150 160 CHARPY V-NOTCH ENERGY YIELD STRENGTH (02% OFFSET.! AT -196 °C (FT-LB) AT -I96°C Fig. 2. The effect of reheating temperature on (a) the Charpy V-notch energy at -196°C and the gra^.n size, and (b) the room tempera ture hardness and the -196°C yield strength, of the Fe-12Ni-0.5Ti alloy. (XBL 731-5659) -134- size is plotted against the reheating tem perature in Fig. 2(a). For lower reheating temperatures in the equilibrium a+Y tempera ture range, some a -*• y transformation took place. However, no grain refinement was observed, Fig. 3. Another observation that can be made from Fig. 3 is the precipitation of Ni3Ti in the a phase. These precipitates were not present in the 700 and 750°C reheated microstructures. Higher reheating temperatures above 7S0°C produced rapid grain growth. Figure 2(b) shows variation of hardness and yield strength with reheating temperature. Note that the 650°C reheating, which produced a structure with precipitate, gave a yield strength 10,000 psi greater than those for the 700 and 7S0°C reheating treatments which produced grain refinement with no precipitation. These initial results suggested that a two-stage heat treatment including grain •efinement in the temperature range 700-750°C and precipitation hardening below 650°C would be beneficial in optimizing both strength and toughness. Figure 3 is an example of this type of microstructure. For specimens with this microstructure, yield strength of 150 ksi, and Charpy impact value of 150 ft-lb were attained at liquid Fig. 3. Transmission electron micrographs N2 temperature. It was believed that the of carbon replicas showing the microstructure usual deterioration of notch toughness from of the Fe-Wt^-O.STi alloy. precipitation aging (grain boundary pre (a) Austen.iized at 900°C and reheated cipitation) was avoided in the present case at 650°C for 2 hr. Charpy V-notch energy by the formation of non-precipitated at -196°C was 10 ft lb. (transformed austenite) regions along the (b) Austenitized at 900°C, reheated at grain boundaries, during the aging step. 700°C for 2 hr and then at 650°C for 2 hr. Charpy V-notch energy at -196°C was 150 ft lb. (XBB 731-378; The above results showed that a fivefold increase in grain refinement resulted through a single cycle into the a ->- y phase region. Past studies on other system? suggest that 6. MECHANICAL PROPERTIES OF THE CRYOGENIC repeated cycling can further refine grain ALLOY Fe-12Ni-0.25Ti size. Initial results on the present system confirmed this view. Hot working in the Sungho Jin and John W, Morris ot+Y two-phase region or cold working fol lowed by direct reheating to the precipitation The excellent combinations, strength and aging temperature range also showed promising toughness of the Fe-12Ni-0.5Ti alloy at microstructures and are thus providing a temperatures down to -196°C have been discus number of processing methods to achieve the sed in an earlier section. The optimum heat desired microstructure. Each of these treatment for this alloy was a 2 hr techniques may be useful for a particular austenitizing treatment followed by a 2 hr industrial application. inneal at 750°C. With this heat treatment, ±e Charpy V-notch energy value at liquid helium temperature (6°K) was good (45-60 Future research activity will concentrate ft lb), though less spectacular than the value on maximizing yield strength and notch tough at -196°C (130-160 ft lb). The function of ness in the manner described above for a given 0.5% Ti in the alloy was to act as a scavenger range of service temperatures. This will en of impurities and also to impart solid tail work on higher Ti concentrations in Fe-Ni solution and precipitation hardening (196 ksi alloys with both higher and lower concentra yield strength at 7°K). tions of Ni than the 124 alloy described here. The principles used in developing the present It was believed that decreasing the Ti Fe-Ni-Ti system will be expanded to include content to 0.251 woulJ considerably improve other promising alloy systems, e.g., the the toughness at 6°K without causing an Fe-Wi-Ti system. -135- appreciable decrease in strength. Tne grain this often results in an increase in creep size was further reduced by a 30? cold work rate. The concept of using intermetallic at room temperature before the final anneal. compounds, rather than carbides, for increas The final treatim-nt of Fe-12 Ni-0.25 Ti alloy ing the high temperature strength of alloys consisted of the following steps: is decades old, out progress has been slow. The class of compounds known as the Laves CI) Forge at 1100°C and air cool. phase have been considered for enhancing (2) Anneal at 700°C for 2 hr and ice brine elevated temperature properties. Other quench. investigators have concluded that, in general, (3) Cold work 304 at room temperature. when the Laves phase is present in significant (4) Anneal at 700°C for 2 hr and ice amounts it can degrade room temperature brine quench. ductility wit'1 little effect on creep properties. The results from the present investigation show, however, that with microstructural control, The 6°K Charpy V-notdi energy was 172 ft lb satisfactory values of room temperature and the yield strength was 185 ksi for the strength and ductility, as well as good specimen heat treated as above. The grain stress-rupture life, can be attained in very size was estimated to be lOu. A two hour low carbon Fe-Ta alloys strengthened with post processing treatment in the 550-6S0°C the Laves phase Fe2Ta. temperature rjnge caused the nucleation of extremely fine (O.Su) grains of the y phase at the boundaries of the existing lOu grains Recently it was demonstrated by Jones of the a phase. On cooling from the 550-650°C et al.i that in iron-rich alloys of the range, the y phase transformed to a, and the Fe-Ta system, the brittleness due to the room temperature microstructure consisted presence of the continuous network of the of 10p grains of a decorated at the boundaries Laves ^hase at the ferrite grain boundaries by 0.5p grains of a. Charpy energy values could be overcome by a simple a •+ y -*• a heat greater than 225 ft lb at 77°K and 220-225 treatment which spheroidized the precipitate. ft lb at 6°K were attained when the post The alloys were then found to possess good processing step was added to the heat treat room temperature ductility and tensile ment. strength. Short time tensile tests at elevated temperatures indicated a gradual Research in progress is aimed at elimina decrease in strength up to 600°C, then a ting the mechanical processing and using sharp decrease as the test temperature was thermal cycling to attain fine grain size. raised to 700°C. The temperatures for the Preliminary grain size studies have indicated a -* y and y -*• <5 phase changes were high in that alternate 700 and 600°C one hour anr.c-.al- binary iron-tantalum alloys, and it was ing treatments with intermediate brine considered desirable to lower these tempera quenches decrease the grain size to the range tures to a range more suitable for heat 0.3-0.7u. The influence of this heat treat treatments by adding a third element to the ment on the inicrostructjre and mechanical alloy. Chromium was chosen for this purpose properties of the Fe-12Ni-0.25Ti alloy is because it is known to enhance also the under investigation. oxidation resistance which is essential for elevated temperature applications. This work was the joint effort of principal investigators Parker, Zackay, and Morris. Investigations of structure and short time tensile properties were carried out on alloys of iron containing 1 at.t tantalum and chromium contents up to 7 at.S. The structure 7. STRUCTURE AND MECHANICAL PROPERTIES OF of the Fe-lTa-7Cr alloy is shown in Fig. 1 fa} NON-CARBON-CONTAINING FERRITIC ALLOYS AT after solution treating at 1320°C (in the 6 ROCM AND ELEVATED TEMPERATURES phase field) for one hour and aging at 700°C (in the a + Fe2Ta phase field) for 40 minutes, Dilip Bhandarkar and Manjeshwar S. Bhat and in Fig. 1(b) after spheriodizing at 1100°C (in the y + Fe2Ta phase field) for 10 minutes Most commercial creep resistant materials following the aging treatment. The derive their strength from a fine dispersion spheroidizing heat treatment not only of a second phase which hinders the movement eliminated the continuous grain boundary of dislocations. In creep resistant ferritic layer of Fe2Ta, but also resulted in grain steels, the fine dispersion almost invariably refinement. Similar changes were observed consists of carbides. The morphology and in the other alloys that were investigated. composition of carbide particles tend to change slowly at elevated temperatures and The short time yield strength is plotted -136- TEST TEMR lOOtfF tSM*C) Fig. 1. Microstructures of the Fe-lTa-7Cr alloy (a) aged at 700°C for 40 minutes and (b) aged at 700°C and spheroidized at 1100°C for 10 minutes. (XBB 7210-4984) 100 RUPTURE TIME, h'RS Fig. 3. Stress vs rupture time at 1000°F (538°C) for the spheroidized Fe-lTa-7Cr alloy. Also shown are the results reported in the literature for the lCr-]Mo-0.25V steel and the 403, 410, and 422 type stain less steels. (XBL '210-7053) 200 400 600 K TEST TEMPERATURE, *C Fig. 2. Yield strength (0.2% offset) as a function of the test temperature for alloys Fe-lTa, Fe-lTa-SCr, and Fe-lTa-7Cr spheroi dized at 1100°C for 10 minutes. (XBL 731-5601) as a function of the test temperature in Fig. 2 for the alloys Fe-lTa/Fe-lTa-5Cr and RUPTURE TIME, HRS Fe-lTa-7Cr that were spheroidized before Fig. 4. Stress vs rupture time at 1100°F testing. The yield strength did not drop (593°C) for the alloys mentioned in Fig. 3. precipitously until a temperature of about Also shown are the preliminary results on 700°C was reached. The Fe-lTa-7Cr alloy the Fe-lTa-7Cr-0.5!4> alloy. was considerably stronger than the Fe-lTa (XBL 7210-7054) alloy. Creep and stress rupture tests were carried out on the Fe-lTa-7Cr alloy at 538°C (1000°F) and 593°C (1100°F). Stress The addition of molybdenum was made to obtain vs time fo rupture data are plotted in Figs. additional solid solution strengthening of 3 and 4. Also shown are the stress-rupture the ferrite. Considerable improvements in properties reported in the literature for stress-rupture properties were achieved as lCr-lMo-0.25V steel and 403, 410, and 422 can be seen from Fig. 4. type stainless steels. Both at 1000 and 1100°F, the Fe-lTa-7Cr alloy results were Research is now in progress to optimize above those reported for the 403 and 410 the heat treating procedure for the alloys stainless steels, but below those for the Fe-lTa-7Cr and Fe-lTa-7Cr-0 jMo. It was 422 stainless steel and the lCr-lMo-0.25V observed that by repeated cycling through steel. Also shown in Fig. 4 are the the a -*• y ->• a phase transformation an preliminary stress-rupture data on a extremely fine grain size was attained with Fe-lTa-7Cr-0.5Mo alloy which was developed a structure similar to a massive martensitic along the same lines as the Fe-lTa-7Cr alloy. structure. It is of interest to investigate -137- whether such a structure is beneficial for The approach is based on the fact that the creep properties. Another area of interest deformation-induced transformation of is to examine the possibility of using the austenite to martensite is strongly dependent Fe2Ta dispersion to pin down the irregular on the stability of the alloy. It was shown grain boundaries resulting from the by Bhandarkar et al.2 that in TRIP steels a •+ y •* a transformation. Preliminary of low stability, a large amount of martensite studies indicate that such a possibility formed on the application of a very low exists, but certain modifications are stress at temperatures above the Ms, but well needed in the heat treating temperatures and below the Md. times. Two TRIP steels were designed to have a low stability by adjusting the chemical 1. R. H. Jones, E. R. Parker, and V. F. compositions which are given in Table I. The Zackay, Use of an Allotropic Phase Change first step in the processing was to deform to Enhance Ductility in Fe-Ta Alloys, Proc. the annealed steels (austenitic) by rolling Fifth Intern. Materials Symp., Berkeley, 5-101 at cryogenic temperatures (-196 and Calif., September 1971, p. 829; -78°C) to form large amounts of martensite. The purpose of the above step was to attain a high dislocation density in the alloy from both the martensite formation and the 8. STRUCTURE AND MECHANICAL PROPERTIES OF deformation. The steels were then heated to TRIP STEELS PROCESSED BY DEFORMATION AND temperatures between 750 and 800°C and held THERMAL CYCLING for short intervals of time to revert the martensite to austenite. Though complete Howard E. Adkins, Jr. and Dilip Bhandarkar reversion of martsnsite was desired, it was not achieved in the present investigation. Metastable austenitic steels with excellent Subsequent repetitions of the process of combinations of strength and fracture tough deformation and thermal cycling were carried ness owe their unique properties to the out on each alloy and the room temperature deformation-induced transformation of austen tensile properties were determined after ite to martensite. Earlier investigations each complete cycle. Several variations of of these steels, known as TRIP steels, have the normal cycling procedure were investigated shown that the mechanical properties are in an attempt to optimize the mechanical strongly dependent on such variables as the properties. Yield strengths up to 140,000 psi chemical composition and the thermomechanical and ultimate tensile strength up to 220.000 processing. >'n important step in the psi were attained. One of the problems processing is the prior deformation of austen- associated with thermal cycling was the ite. This consists of deforming the austenit carbide precipitation which caused the ic steel up tc an 80? reduction in thickness depletion of the austenite stabilizing ele by rolling above the Mj, which is the tempera ments from the matrix. On testing at room ture above which no martensite can be produced temperature, the unstable austenite in the by deformation of the austenite. The primary steel transformed to martensite at very low purpose of the prior deformation is to stresses and the transformation caused gross increase the yield strength of austenite. yielding of the tensile specimen. Several specimens tested at 100°C exhibited a higher Recently, Koppenaal* showed that an in yield strength than that at room temperature. crease in the strength of austenite could At 100°C the austenite was more stable than be obtained by-thermally cycling the alloy at room temperature and a higher stress was required to initiate its transformation to between a temperature within the Ms - Mf range and the temperature at which martensite martensite. reverts to austenite by reverse shear. The Table I. Chemical compositions of steels. reversion results in a change in crystal structure from the body centered tetragonal to the face centered cubic while maintaining the high dislocation density associated with Alloy the martensite structure. Koppenaal concluded Compositions (wtl) that the process could be conveniently designation applied only to steels which had their Ms C Cr Ni Mn Mo temperature above -196°C. A 0.28 12.0 7.8 1.9 The present investigation deals with the development of an alternate approach to B 0.30 9.0 7.8 2.0 2.0 apply the process of reversion to TRIP steels, some of which have their Ms below -196°C. -138- 1200 1. T. J. Koppenaal, A Thermal Processing Technique for TRIP Steels, Met. Trans. 3, 1549 (1972). 2. D. Bhandarkar, V. F. Zackay, and E. R. Parker, Stability and Mechanical Properties of Some Metastable Austenitic Steels, Met, 900- Trans. 3, 2619 (1972). UJ 9. DESIGN OF HIGH MANGANESE METASTABLE IE AUSTENITIC STEELS 3 fc 600- David G. Atteridge A major stumbling block to increased industrial use of TRIP steels is the high cost of alloying elements and the processing. I A typical TRIP steel contains 9% Cr, 8* Ni, 3* Mo, 2% Ml, and 0.31 C. Nickel and 300- molybdenum are particularly expensive. Typical processing requires an 80% reduction in thickness by warm rolling at 45CC and involves high-strength rolls and high load capacity mills. The present investigation deals with the appliescion of alloy theory and use of the TRIP phenomenon to develop steels with substantially reduced cost from 0.4 0.8 1.2 .6 the standpoint of both alloying and proces CARBON CONTENT, WT. PCT sing. Fig. 1. Vertical section of the ternary A series of Fe-12% Mn alloys with carbon diagram of the Fe-12% Mn-x% C system. contents from zero to 1.3% were selected for (XBL 731-5646) the present investigation. The equilibrium phase diagram of the Fe-12% Mn-x% C system raised. The rate of formation of martensite predicts structures that are multiphase with strain during testing increased with containing ferrite, austenite, and carbide an increase in rolling temperature, while in varying proportions, depending on the the volume fraction of martensite at fracture carbon content, as shown in Fig. 1. The decreased. austenite in certain alloys of this system is metastable and undergoes a strain-induced The 0.2% C alloy was magnetic at room transformation to body centered tetragonal temperature when slowly cooled from the martensite. Following an austenitizing austenitic region. Deformation by rolling treatment, the alloys were deformed by in the austenitic region, however, mechanical rolling at several different temperatures. ly stabilized the alloy and the deformed The role of the deformation on structure of alloy remained austenitic at room temperature. alloys was twofold. First, deformation of The austenite was unstable and on testing at an alloy in a phase field where part of the room temperature it exhibited a stress- structure was austenitic resulted in a induced martensitic transformation. This partial transformation of the austenite to resulted in a low vield strength and a high body centered tetragonal martensite. Second ultimate strength. ly, chemical changes, in the form of carbide precipitation, occurred in carbon-containing A comparison of the yield strengths of alloys and these in turn caused a decrease in alloys with different carbon contents, rolled stability of the alloy as a result of depletion at the same temperature, showed an increase of alloying elements. in yield strength as the carbon content increased. The percent elongation was the The carbonless Fe-12% Mn alloy shewed a largest for metastatic austenitic alloys that continuous decrease in yield strength and were produced by rolling in the austenite ultimate strengths and an increase in the per phase field and smallest for alloys rolled cent elongation as the rolling temperature was in the three-phase austenite-ferrite-carbide region. -139- Based on the mechanical properties and the transformation behavior that were observed, a two-stage rolling schedule was designed. The alloy was first rolled 624 at either 950 or 750°C and then at temperatures ranging from ISO to 550°C until a cumulative reduction of 804 was attained. For the 0.24 C alloy rolled in two stages at 750 and 400°C, a yield strength of 165,000 psi, an ultimate strength of 230,000 psi, and an elongation of 174 were observed. The higher rolling temperature (at 750°C) results, of course, in lower roll separation forces and is therefore a desirable processing feature. The results were encouraging and further research is in progress. 10. ISOTHEHMAJ STUDIES ?« PHASE TRANSFORMATIONS IN STEELS IS 30 60 120 600 3600 7200 B. Naga Prakash Babu Tlmi ,t fMtl The conventional methods of plotting an T—i 1—i 1 r I-T diagram are characterized by poor accuracy, are extremely slow and tedious, and require the metallographic examination of a series of samples reacted for various times. Also, it is possible that the interruption, by quenching, of the reaction for examination of structure would change the kinetics. An alternative approach is to observe the change of some property associated with a phase change in the sample. The method presented here consists of continuously recording the change in the mag netic susceptibility of the sample under going austenite transformation. For a given alloy, the magnetic susceptibility is a Fig. 1. (a) Percentage of austenite function of temperature, strain, and structure isothermally transformed as estimated from the of the phases present. magnetic susceptibility change. (b) The I-T diagram of AISI 4340 steel The apparatus consisted of an inductor plotted from the data in Fig. 1(a). coil, 5 in. long and 2 in. in internal (XBL 761-5644) diameter, oscillating at 1.4 kHz, and generating a weak field around the sample. The coi"' surrounded a 1 in. outside diameter punch recorded the austenite decomposition stainless steel tube, into which a sample continuously at intervals of 0.77 sec, strip of dimensions 3 in. x 1/2 in. x 0.03 in. 7.7 sec, or 77 sec. To interpret the data was quenched. To maintain isothermal quantitatively, a calibration was needed for conditions, and to obtain a rapid quench, the the temperature dependence of the magnetic quenching medium was circulated through the susceptibility. For this reason the data tube by pumping it from a heated reservoir. obtained at different temperatures were The sample was austenitized ti'.« dropped into normalized to a single temperature. the coil zone of the tube. The change in the magnetic susceptibility of the sample Research is in progress to use the caused a change in the inductance of the technique to study the bainite and martensite coil, which, in tum, changed the frequency transformations in a commercial AISI 4340 of a reference oscillator. The period of steel. The results presented here are oscillation was measured by a time interval tentative, because the curves of percent counter, a crystal oscillator providing a austenite transformed as estimated from the constant time base. The digital output from isothermal susceptibility data obtained at the counter was fed into a tape punch. The different temperatures were only approximately normalized to room temperature. Figure 1(a) Materials--Techniques and Applications of shows the various isothermal curves and Klectron Microscopy, University of California, Fig. 1(b) gives the I-T diagram derived from Berkeley, September 13-17, 1971, edited by Fig. 1(a). G. Thomas (university of California Press, Berkeley, 1972), p. 798. The martensitic reaction in this low alloy steel was believed to be athermal. The 2. R. H. Jones, E. R. Parker, and V. F. bainite nose of the I-T diagram for AISI 4340 Zackay, "Use of an Allotropic Phase Change shown in Fig. 1(b) is very much the same as to Enhance Ductility in Fe-Ta Alloys," ibid., the one in I-T diagrams determined by other p. 829. investigators. Between 2S0 and 300°C, it is believed that there was a martensitic 3. W. W. Gerberich and V. F. Zackay, The reaction in the first 15 seconds, and then Potential of Metastable High-Strength Fibers the reaction proceeded by a bainitic mode. for Strong, Tough Composites, Met. Trans. Also, the bainitic reaction was complete 3. 747 (1972). within an hour between temperatures 300 and 400°C. Above 400°C the bainitic transforma 4. V. F. Zackay, E. R. Parker, R. D. tion was incomplete. At 490°C there appeared Goolsby, and W. E. Wood, Untempered Ultra to be an onset of the pearlitic node of High Strength Steels of High Fracture Tough transformation, after partial bainitic ness, Nature Physical Science 236, 108 (1972). transformation. 5. D. Bhandarkar, V. F. Zackay, and E. R. Parker, Stability and Mechanical Properties cf Some Metastable Austenitic Steels, Met. 11. RESEARCH PLANS FOR CALENDAR YEAR 1973 Trans. Z, 2619 (1972). Victor F. Zackay and Earl R. Parker 6. R. H. Jones, V. F. Zackay, and E. R. Parker, Laves Phase Precipitation in Fe-Ta Further studies of the influence of Alloys, Met. Trans. 3_, 2835 (1972). microstructure and composition on the fracture toughness and fatigue crack propagation of Papers presented ultra-high strength steels are planned for the coming year. In particular, the effect 1. E. R. Parker and V. F. Zackay, "Enhance of impurities and inclusions on the fracture ment of Fracture Toughness in High Strength toughness of AISI 4130 and AISI 4340 will be Steel by Microstructural Control," Symposium investigated. Studies of the tempering on Fracture and Fatigue, George Washington behavior of secondary hardening steels will University, Washington, D. C., May 3-5, 1972. be continued. Additional studies of the (LBL-832) phase transformations in cryogenic iron- nickel-titanium alloys will be made in order 2. V. F. Zackay, E. R. Parker, and to get microstructures that a*e beneficial D. Bhandarkar, "The Structure and Properties for high toughness and strength at tempera of a Non-Carbon Containing BCC Iron Alloy tures down to that of liquid helium. A new at Room and Elevated Temperatures," Proc. project will be initiated with an objective The John E. Dorn Memorial Symposium on Rate of designing iron-manganese-titanium alloys Processes in Plastic Deformation, Cleveland, for cryogenic applications. The research Ohio, October 16, 1972. (LBL-1178) on elevated temperature alloys will involve further optimization of composition and LBL reports microstructure in order to improve the stress rupture properties of Fe-Ta-Cr-Mo alloys. 1. C. J. Bruggeman, The Effect of One- Dimensional Plane Strain Waves in a-Phase Titanium-Aluminum (Ph. D. thesis), LBL-802, April 197Z. 12. 1972 PUBLICATIONS AND REPORTS 2. A. J. Baghdasarian, Corrosion Resistance Victor F. Zackay, Earl R. Parker, and of TRIP Steels (M. S. thesis), LBL-806, Associates June 1972. Journals and books 3. M. W. Perra, Nuclear Magnetic Resonance Study of GP Zone Formation in Aluminum -41 1. R. D. Goolsby, W. E. Wood, E. R. Parker, Copper (M. S. thesis), LBL-838, June 1972. and V. F. Zackay, "Effect of Lath Boundary Precipitation on Fracture Toughness of Martensite," The Structure and Properties of -141- 4. R. A. McCoy and W. W. Gerberich, 5. W. A. Hoiwood, Factors Controlling the Hydrogen Bnbrittlemmt Studies of a TRrP Fracture Toughness of Cryogenic Alloys Steel, LBL-439, July 1972 (accepted by (M. S. thesis), LBL-1121, Dec. 1972. Met. Trans.). -142- F. HIGH FIELD SUPERCONDUCTIVITY Earl R. Parker, Viator F. Zaakay, and Milton R. HokuB, Principal Investigators 1. INTRODUCTION A composite superconducting tape, consist ing of thin filaments of Nb3Sn dispersed in a niobium matrix, has been developed at MRD. Much of the current work was directed towards improving the electrical properties of the tape. During the past year a sixfold increase in current-carrying capacity was achieved. Several means were found for increasing the volume fraction of Nb3Sn in the composite. Present work on enhancing the critical current density of the Nb3Sn filaments should lead to further inprovement. Niobium powdeT is available from commercial suppliers in a rather limited range of particle size distri butions. In an investigation of the possibil ity of using niobium hydride in place of elemental niobium powder, it was demonstrated that this substitution is entirely practi cable. Furthermore, it was found that from the hydride any desired particle size range could be prepared here in the laboratory, as illustrated in Fig. 1. This of course provides much greater flexibility in optimizing the process variables. 1. M. R. Pickus, V. F. Zackay, E. R. Parker and J. T. Holthuis, A Powder Rolled Super conducting Composite with Controlled Morphology, LBL-821 preprint, August 1972. (scheduled for publication in the Interna tional Journal of Powder Metallurgy, January 1973.) Fig. 1. Scanning electron micrographs of niobium hydride powder. Cx 1000) (a) As received from the supplier. (b) Ball-milled 48 hours and treated with hot hydrochloric acid. (XBB 731-371 % XBB 731-370) 2. EVALUATION OF OPTIMAL PARAMETERS FOR POWER-ROLLED TAPE Kanithi Hemachalam A few illustrative results of this investi gation are presented here. The ductility of the porous sintered niobium tape is an important consideration for two reasons: 1) The infiltrated tape must withstand a cold reduction in thickness of the order of 801 in order to impart to the infiltrated tin a filamentary morphology, and 2) The diffusion reacted tape must with stand the winding required to fora solenoids. Samples of green tape were sintered for various times at several temperatures. Because of the physical dimensions of the tape, a standard tensile test for determining the percentage elongation was considered unsuitable. Instead, a somewhat crude but effective method was employed. A sintered specimen was bent around rods of varying diameter, beginning with the largest diameter. The diameter around which the bending resulted in a visible crack was designated as "The Critical Diameter." The relationship be tween critical diameter and sintering conditions is shown in Fig. 1. As pointed out by Prakash Babu (LBL-437}, there is a progressive reduction in the volume fraction of tin with increasing amounts firttn Sfript Length 49 in. Width 0.41". Thicknm OOI9in. 1.0 Poroilty 2fl±l% Furrwe* Pff isura; Inlllot I.I itIO"*mm Hg Maximum 2.0*10-* mm Hg .£' 0.8 f 5 0.6 *o. 0.2 2200*0 0 0 12 3 4 SinterinQ Time, fflin. Fig. 1. Critical diameter (an indication of ductility) as a function of time at various temperatures. (XBL 7210-70441 -144- l—i—r "i—I—i—I—r PARALLEL FIELD IMRD &E.(22CY0325) = 1000 600 400 zoo 100 40 60 100 KG FIELD Fig. 3. Comparison of current-carrying capacity (dc) as a function of applied magnetic field for IMRD and General Electric Co. type 22CY0325 superconducting tapes. (XEL 731-5626) of mechanical deformation. He found that tin from the pores had squeezed out to the surface during cold reduction, due to the dif ference in flow stresses of niobium and tin. Such losses, of course, have an adverse effect on the current-carrying capacity of the fully processed tape. Several pi-ocedures for preventing this loss of tin were devised and investigated. A marked improvement was observed with the following methods: 1) Cladding the infiltrated tape with Fig. 2. Influence on volume fraction of niobium or copper before rolling. Nb3Sn of various methods of constraining 2) Performing the rolling cryogenically. tin loss due to plastic flow. (x500) 3) Hardening the infiltrated tin by Filaments: Nb3Sn; Matrix: Niobium; alloying with copper. White: Unreacted tin. Examples of the increase that has been (a) Normal processing: rolling at room achieved in the volume fraction of M>3Sn in temperature. the fully processed tape are shown in Fig. 2, (b) Using a dilute (1-2%) solution of for a reduction in thickness of 80%. copper in tin as the infiltrant. Figure 2(a) shows the tape after normal (c) Rolling the infiltrated tape at -78°C. processing--rolling at room temperature with (d) Cladding the infiltrated tape with no special treatment. The sample shown in niobium before rolling. Fig. 2(b) was prepared by using a dilute er (XBB #'s 731-369, 368, 367, 366 respectively) solution of copp in tin for the infiltrant. The copper content was between 1 and 2% by -145- weight. For Fig. 2(c) the Infiltrant was pure tin, but the reduction in thickness was carried out at -78°C. Cladding the infil trated tape with niobiim before the rolling Nb3Sn was performed resulted in the structure shown in Fig. 2(d). Other topics investigated and described in LBL-1108 include: the strength of the tape at sintering temperatures, a comparison of powder flow during rolling in air and vacuim, the influence of 60 cycle vibration on powder flow, along with a number of design considerations relevant to the preparation of continuous tape. On the basis of the limited data presently available, a comparison of the critical current-applied magnetic field relcitionship for the newly developed, undoped, IMRD tapo and General Electric Co.'s doped tape, one of the best of the commercially available tapes, is shown in Fig. 3. At low fields the influence of doping in increasing the critical current density of NbiSn appears to be the dominant factor. As the strength of Fig. 1. Microstructure of IMRD superconduct the magnetic field is increased the unique ing tape showing Nb3Sn filaments interfaced filamentary structure of the IMRD tape, even with copper. (xlOOO) (XBB 731-11) though undoped, rapidly exhibits its advantage as a high field superconductor, the cross over occurring at about 55 kG. It is previously noted, used small additions of anticipated that the advantage would be even copper for alloy hardening of the infiltrated more pronounced under ac conditions. Doping tin in order to minimize its loss during the IMRD tape, it is hoped, will retain the rolling, the use of 25 at.% copper so slope of the curve, but shift it upward to hardened the tin that it became necessary to higher current values. perform the rolling at elevated temperatures in order to obtain a filamentary micro- structure. The rolling temperature of 480°C was selected on the basis of such considera 3. A NEW PROCESS FOR MAKING STABILIZED tions as the formation of undesired SUPERCONDUCTING TAPE intermediate phases, optimum tin retention, and development of the desired microstructure. William H. Harris A diffusion-reacted composite consisting of a niobium matrix containing copper-interfaced The powder metallurgy irethod for making filaments of Nb3Sn is shown in Fig. 1. IMRD superconducting tape was modified to produce Nb3Sn filaments interfaced with cop Pulsed magnet tests to determine the per. It is believed that such a micro- critical current-critical field relationship structure could desensitize NbjSn with are presently under way for samples with a respect to instabilities such as short-term 75, 85, and 95% reduction in thickness. flux jumping. During such instabilities the Magnetic fields up to 200 kG will be employed. copper interface would provide a low resis tivity path for supercritical currents. This would have the effect of increasing the crit ical current density of the Nb3Sn filaments. 4. THE NIOBIUM-MOLYBDENUM NITRIDE SYSTEM For this purpose, the infiltrant employed was an alloy of 75 at.S tin and 25 at.% Michael M. Collver, William H. Gourdin, copper. The idea w'as that the tin would Charles W. Krause and Steve Wilson diffusion-react with the niobium, leaving the copper behind as an interface. With Some recently developed theoretical ideas this alloy maintained at a temperature of indicate that the critical temperature, Tc, 655°C, infiltration of the porous niobium of niobium nitride should be increased by was complete. Whereas Hemachalam, as substituting molybdenum for some of the 146- niobium. A study of the ternary up to 15% process leads to some cross-linking of the molybdenum was therefore undertaken. The filaments--which is presumed to be undesirable compounds were prepared by flowing nitrogen for ac superconductors. The specific objec gas taken from above the liquid in a liquid tive would be to develop a filamentary nitrogen dewar over foils of the alloys composite with no cross-linking. under the following conditions: 10 minutes and 30 minutes at 1400°C and 30 minutes at The increase in current-carrying capacity 1550°C. The data indicate, rather than an already achieved has made necessary a increase, a progressive decrease in Tc, modification of the current-source system amounting to approximately 1.5°K for 15% of the pulsed magnet apparatus. The changes moi/bdenum. to make more amperes available will be ef fected as soon as possible, so that evaluation of both the niobium-tin-copper and the zirco nium-doped tapes can be carried out. 5. RESEARCH PLANS FOR CALENDAR YEAR 1973 As progress is made in the development of Harl R. Parker, Victor F. Zackay, and ac superconductors, provisions will have to Milton R. Pickus be made for the performance of ac tests. Since data on the composites will be available Research will continue on optimization of from dc and pulsed magnet tests as well, an IMRD superconducting tape. It has been attempt should be made to establish the reported that doping with certain elements, extent and validity of correlations between particularly zirconium, increases the current- the various types of data, particularly with carrying capacity of niobium-tin super pulsed data as a reference. conductors by a factor of 3-4. An investiga tion will be made to determine whether a comparable enhancement can be obtained for powder-rolled tapes. Two approaches will be used: zirconium will be added to the infil- 6. 1972 PUBLICATIONS AND REPORTS trant tin bath, in one instance, and, in the other, a pre-alloyed niobium-zirconium "arl R. Parker, Victor F. Zackay, powder will be used instead of the elemental Milton R. Pickus, and Associates niobium powder which is presently employed. The influence of zircon* = in on infiltration LBL reports and diffusion kinetics will be determined for both approaches, in addition to its 1. M. R. Pickus, V. F. Zackay, E. R. Parker, effect on current-carrying capacity. and J. T. Holthuis, A Powder Rolled Super conducting Composite with Controlled Morphol Feasibility experiments are contemplated ogy, LBL-821 August 1972 (submitted to utilizing different compaction modes with Int' 1 Journal of Powder Metallurgy). the objective of designing a superconducting material better suited for ac applications. 2. K. Hemachalam, Evaluation of Optimal The present tape is a composite consisting Parameters for the Production of Superconduct of Nb3Sn filaments dispersed in a niobium ing Niobium-Tin Tape in Relation to Pilot ma'crix. The filamentary structure is Plant Design (M.S. thesis), 1BL-1108, June desirable. Unavoidably, however, the present 1972. -147- G. POWDER METALLURGY Milton R. V-iakuSj Principal Investigator b. Alloy Design for Powder Metallurgy 1. INTRODUCTION The initial effort will center on ferrous systems, because of their technological Milton R. Pickus importance. It is generally considered, with much supporting evidence, that the IMRD research in powder metallurgy (P/M), deficiency in mechanical properties noted which in the past was limited to superconduc in the introduction is attributable to ting materials, is being broadened to residual porosity. Most commercially encompass engineering alloy systems of a produced P/M parts have an unintentional more generalized nature. Although P/M has residual porosity of the order cf 10% or become an important technology—from it have more. In this range there is a substantial come, in addition to its intrinsic economies, reduction in mechanical property values. such significant developments as the tungsten filaments for the incandescent lamp, the Two basic approaches to the problem of sintered carbides for cutting tools and residual porosity have been critically wear-resistant dies, the self-lubricating examined: bearing, the friction brakes for modem 1. Design alloy systems that are conducive aircraft, to name but a few--its further to the attaijiment of near theoretical density. development depends upon a systematic 2. Design alloy systems that have a marked application of the principles of materials insensitivity to the presence of residual science. As compared with wrought materials, pores. structural P/M products made by pressing and sintering are deficient in important mechani A study of the available data as reported cal properties such as ductility, fatigue in the literature indicates that enhancement strength and fracture toughness. To improve of fracture toughness should he regarded as these properties current practice is to an expedient rather than a basic solution. resort to secondary treatments which are In those instances where fracture toughness, costly and not always completely effective. at a given level of residual porosity, has There is a good possibility that this been increased by alloying, it has been found situation is simply a consequence of applying that the fracture toughness of the alloy P/M techniques to the standard engineering rapidly increases still further as the alloys made by conventional processes with porosity is decreased. Moreover, such little, if any, modification. The objective fundamental material properties as Poisson's of this aspect of research in P/M is to ratio and the modulus of elasticity, while design alloy systems that are especially insensitive to changes in composition or adapted to the sequence of powder compaction thermal treatment, are directly proportional and sintering. to density. Tensile strength, yield strength, hardness, and elongation are all nearly proportional to density. Finally, impact strength appears to fall off exponen 2. RESEARCH PWNS FOR CALENDAR YEAR 1973 tially with decreasing density. Milton R. Pickus It has been concluded, therefore, that the significant solution for the problem of a. Superconducting Materials by Powder residual porosity would be its elimination, Metallurgy and that investigating the means for achieving full density would be the focal Research will continue on optimization of point of this research effort. the powder-rolled superconducting tape developed at IMRD. Whereas the progress To implement this objective, several and research plans for superconducting jiEthodologies will be employed. These materials are described in another part include liquid phase sintering, dispersions of this report under the title, of selected phases, doping, and the use of "High Field Superconductivity," it may be very fine powders. briefly noted here that a sixfold increase in current-carrying capacity of the tape As progress is made with simple alloy has been achieved during the past year. In systems the results will be applied to more the coming year feasibility experiments are complex multi-element systems ''"ith emphasis contemplated to develop a superconducting on high strength and corrosion-resistant material especially suited for ac alloys. applications. -148- H. THERMODYNAMICS OF METAL SYSTEMS Ralph R. Hultgrent Vvinovgal- Investigator 3. THERMODYNAMIC EVALUATION AND COMPILATION OF METAL ALLOY SYSTEMS 1. EVALUATION OF THESM3DYNAMIC DATA Leo Brewer and Ralph R. Hultgren Ralph R. Hultgren and Pramod D. Desai The bibliography of the literature pertain The evaluation of thermodynamic data for ing to phase diagrams and thermodynamic data metallic systems was completed. The results for metal alloy systems will be kept up to are to be published in two volumes by the date, and evaluations of selected binary and American Society for Metals, as follows: ternary systems will be published periodically. I. Selected Values of the Thermodynamic Properties ot the Elements (.636 pp.,), by Ralph R. Hultgren, Pramod D. Desai, Donald 4. 1972 PUBLICATIONS AND REPORTS T. Hawkins, Molly Gleiser, Kenneth K. Kelley, and Donald D. Wagman. Ralph R. Hultgren and Associates Nonmetallic elements in this volume were In addition to the books mentioned previous evaluated by Donald D. Wagman of the National ly, the following publications were produced. Bureau of Standards. This volume is now in press. Journals II. Selected Values of the Thermodynamic 1. H. I. Yoon and R. R. Hultgren, The Effect Properties of Binary Alloys (143!> pp.), by of Ordering on Lattice Heat Capacities of Ralph R. Hultgren, Pramod D. Desai, Donald Ordered and Disordered AuCuj, J. Chem. T. Hawkins, Molly Gleiser, and Kenneth K. Thermodynamics ±, 375-380 (1972). Kelley. 2. H. I. Yoon and R. R. Hultgren, Heats This volume is now in press. of Formation of Solid Indium-Lead Alloys, J. Chem. Eng. Data 17, 176-177 (1972). 3. J. Chipman, Thermodynamics and Phase 2. METALS HANDBOOK PHASE DIAGRAMS Diagram of the Iron-Carbon System, Met. Trans. 3, 55 (1972). Donald T. Hawkins and Ralph R. Hultgren LBL Reports Approximately 400 phase diagrams were edited and corrected for the forthcoming edition of 1. J. C. Cases, The Heat Content of NaN07 Metals Handbook by the American Society for Between 450 and 800 K, LBL-842, May 1972. Metals. o± these, approximately 300 were taken from Selected Values of Thermodynamic Proper 2. J. C. Cases, Diphenyl Ether, A Versatile ties of Binary Alloys. Substance for Laboratory Demonstrations, LBL-868, June 1972. I. MATERIALS Robert H. Bragg, Principal Investigator 3. SMALL-ANGLE SCATTERING BY ORIENTED ELLIPSOIDS 1. STRUCTURE AND PROPERTIES OF SOME ADVANCED MATERIALS Faysal Hamzeh, Madan M. Biswal, and Robert H. Bragg Robert H. Bragg Small-angle scattering experinsnts reported We have been studying the structure and in UCRL-20517 confirmed certain aspects of properties of Glassy Carbon, GC, and earlier results obtained on pyrolytic graphite Pyrolytic Graphite, PG, the former having by Bragg and co-workers. More careful properties totally different from conventional experiments, over a larger angular range for carbons and graphites, and the latter being which the intensity varies by a factor of similar to pure graphites. During the past 10*, confirmed the volume effect reported year we have done a great deal of work, previously and revealed the expected especially on pyrolytic graphite, and the deviation from the h'3 dependence of the combination of careful experiments and good intensity at large h [h » (4uA)sin 9], theoretical work have produced a greatly predicted by Porod, but left unexplained the enhanced and simplified technique of significance of S in the Porod formula for structural characterization We have also oriented particles, even after corrections for continued our studies of unidirectional local density fluctuations are made. New solidification of eutectics in order to theoretical work reported in LBL-1176 and produce unique microstructures with hereto further work in progress provides a simple fore unobtainable properties. To date our interpretation of both the Guinier and Porod experiments have been confined to the regions of the scattering curve, and enables Al-CuAl2 pseudo-binary eutectic, and signi an unambiguous determination of the size and ficant progress has been made in identifying shape of the particles (voids in the case of the factors governing eutectic solidification PG) for ellipsoids of revolution of arbitrary at high growth rates. shape. Additional new work now being written for publication sxtends the theory to general Excessive delays in the delivery of our ellipsoids of semi-major axes a, b, c, all low temperature electrical measurement different, and the results reduce, for appro- apparatus have occurred, but compensating piate special cases, to those obtained for developments have rendered these delays ellipsoids of revolution. Moreover, these irritating but not crippling. Significant results give a simple proof of Bragg's rule comments are given in short reports given for small-angle scattering which had pre below. viously been established empirically. 2. STRUCTURE OF GLASSY CARBON 4. UNIDIRECTIONAL SOLIDIFICATION OF Al-CuAl2 EUTECTIC Madan M. Biswal and Robert H. Bragg Kwame Ankra and Robert H. Bragg Work to date has shown that heat treatment of glassy carbon previously carbonized at In this work, LBL-1182, banded micro- 1000°C for one hour produces no perceptible structures were studied in the Al-CuAl2 changes in structure over long periods of eutectic, unidirectionally solidified at time at temperatures below 2500 C. Heat large velocities (ca 200 cm/hr) in a horizon treatments at 2700°C have been made but no tal direction in an open mold. Some new analyses have been completed. Mo new types of banding (see Fig. 1} were observed evidence has been found which contradicts including a laminar-to-dendrite breakdown, the inter wound spaghetti model for the pores (see Fig. 2) Some of the observations could in Glassy Carbon, but the heat treatments not be explained by current theories or must be extended to at least 3000°C to possible experimental errors. Analysis of complete the investigation. the ndcrostructures has eliminated all of -150- the known theories, and a mechanism, based on correctional currents in the melt, is proposed but awaits experimental verification. This work was the M.S. thesis of Kwame Ankra, and is being written up, slightly revised, for publication. 5. OTHER TORK IN PROGRESS Robert H. Bragg Louis Salmon has completed his preliminary determination of the experimental conditions necessary to obtain well-aligned Al-CuAl2 composites with arbitrarily chosen inter- lamellar spacings. Ram Saxena is investiga ting a recently observed acousto-electric effect in pyrolytic graphite in which pres sure applied in the c-direction causes appreciable changes in the electrical con *;'',$j«liil!j ductivity in the same direction. The results thus far are preliminary, and the exact mechanism has yet to be established. Madan Biswal has nearly completed his studies of heat treated glassy carbon. Fig. 1. Typical banded structure visible to the unaided eye. Mag. 45* Growth direction is from bottom to top. (XBB 7210-5211) 6. RESEARCH PLANS FOR CALENDAR YEAR 1973 Robert H. Bragg Completion of the work on the graphitiza- tion of glassy carbon is expected by the end of the spring quarter. Also, it is expected that the low temperature facility will be in operation by the end of the winter quarter and the experimental work on the low tempera ture electronic properties of glassy carbon and other materials including extremely high purity Ge can begin. Material available from the solidification studies will also be studied at low temperatures, where unusual changes in conductivity are expected. Work on the acousto-electric effect in pyrolytic graphite will be extended, and studies of the electron-wind effect in Al-CuAl2 eutectics will be initiated. The Al-Si eutectic appears from the phase diagram to be relative ly simple to grow and should have interesting electronic properties, and experiments on the growth and electronic properties of this material will be initiated. 7. 1972 PUBLICATIONS AND REPORTS Robert H. Bragg and Associates Fig. 2. Longitudinal micrograph showing a eutectic-dendritic breakdown at a band. Published paper Growth direction is from bottom to top. Mag. 1420* (XBC 726-3484) 1. F. M. Hamzeh, Formation and Destruction -151- of Magnetic Islands in Torroidal Systems, 2. Kwame Arikra, Unidirectional Solidification (LBL-1143, October 1972) Phys. Rev. Letters of Al-CuAl2 Eutectic (M.S. thesis), LBL-1182, 29, 1492 (1972). December 1972. LBL reports 3. M. M. Biswal and R. H. Bragg, Optics of Bonse-Hart Apparatus, LBL-80S, Feburary 1. F. M. Hamzeh and R. H. Bragg, Small Angle 1972. Scattering of X-rays from Oriented Ellipsoids of Revolution, LBL-1176, October 1972. o, 2 III. Ceramic Science 0 ..&:, -155- A. HIGH TEMPERATURE REACTIONS Alan W, Searcy, Principal Investigator 2. THE FRAGMENTATION OF DIMERS OF METAL TRIHALIDES BY ELECTRON IMPACT 1. THE KINETICS OF VAPORIZATION OF BAMUM SULFATE Alfred Biichler, James A. Roberts, Jr., and Alan W. Searcy Pirooz Mohazzabi and Alan W. Searcy In the report of their investigation of Equilibrium vapor pressures of barium the vapor over LaF3(s) Skinner and Searcyl sulfate and free surface sublimation pressures concluded that the monomer/dimer pressure of (001) face of its single crystals were ratios calculated by conventional analysis both measured for the first time from a phase of their mass spectrometric data were transformation temperature at 1422 to 1540°K, probably too high. They reasoned that this using Torsion-effusion and Torsion-Langmuir was caused by extensive fragmentation of the techniques, respectively. The reaction was dimer molecules under electron impact in confirmed as the ion source of the mass spectrometer to fragments that were masked by fragments of BaS04(s) = BaO(s) + S02(g) + 1/2 02(g) the monomer molecules, which were present at higher concentrations. The same possibility for both equilibrium and free surface con was Taised in the analysis of data subse ditions. Calculations by the Second Law quently gathered on the vapor over CeF3(s).2 method yielded, for the reaction, To gain an indication of whether or not AH° = 138.9 ± 1.9 kcal, massive fragmentation of these dimers is likely we have examined the fragmentation pattern of the vapor over AICI3(s) at a temperature (350°K) where the dimer is known AS° = 60.2 1.3 eu , p ± to be the major species CPAlCl3/ Al2Cl6 < 10~3).5 The major fragment observed, for the equilibrium studies, and AI2CI5 (analogous fragments were attributed to the diners in the CeFj and LaF3 work), has an intensity greater than six times AH*. = 136.4 ± 2.9 kcal, that of any other fragment, such as A1C12*, that could be attributed to the monomer. AS* = 48.0 * 2.0 eu , Thus fragmentation of AI2CI6 to ions that could be masked by the monomer is not a major reaction. for the free surface sublimation studies.''' Thus, within the experimental error, This result does not completely rule out AH*. = AfP but AS* < AS° . extensive fragmentation for the lanthanide element fluorides, but it does make it less likely, since fragmentation patterns for The rate of the free surface decomposition chemically similar molecules are generally reaction was constant, within the experimen qualitatively similar. The anomalous tal scatter in data, during the period of the behavior of La2F6 an^ Ce2F6 appears to be due measurements. This result indicates that to selective scattering from the beam outside effusion of the product gases through the the effusion cell2»4--a previously un porous product lr;er is not rate limiting recognized source of error in mass spectrom even when the pi ..duct layer thickness is as eter studies. With the source of error great as 1 mm. The simplest model consistent identified, the monomer/dimer pressure ratios with these facts assumes that only 3J of for the LaF3 and CeF3 systems can be cal the surface sites are active for the vaporiz- culated by conventional means. tion reaction. 1. H. B. Skinner and A. W. Searcy, J. Pbys. AHy, AS? are the standard enthalpy and Chem. 75, 108 (1971). entropytfor the equilibrium reaction while 2. J."X Roberts, Jr. and A. W. Searcy, AH*,, ASv are called the apparent activation High Temp. Sci. 4, 411 (1972). enthalpy and entropy for the vaporization 3. JANAF Thempchemical Tables, 2nd ed. reaction. NSRDS-NBS 37 (1971). 4. D. J. Meschi, J. Phys. Chem. 76_, 2947 (1972). See following abstract. -156- 3. THE MECHANISM OF ZnO SINGLE-CRYSTAL where U^B is the ratio of molecular weights VAPORIZATION MA/MB- This approximation results in too low a value for *AB, so that its substitution Peter P. Bihuniak, David J. Meschi and in Eq. (1) gives an upper limit for S^g. Alan W. Searcy Application of Eq. (1) to typical examples The relative abundance of vapor species taken from high temperature mass spectroscopy2 from the reaction ZnOj-s)*Zn(g) + 1/2 02(g) reveals that at high source pressures an in their ground and first excited electronic appreciable fraction of some species may be states were compared for equilibriun condi lost from the beam--as much as a third in tions at 148S°K and Langmuir free surface one example.2 The fraction lost will vary vaporization conditions at 1640°K by measuring with the species because of the parameters deflections of vaoor beams in an inhomogeneous PB> °AB, aiui *AB- In general, larger magnetic field, Vh.e fractional occupancy of molecules and minor species tend to be scat zinc atoms in the first excited 3p state was tered most. found to be approximately the same under both Khudsen and Langmuir vaporization conditions. It was not possible to obtain meaningful results for the oxygen vapor 1. D. J. Meschi, J. Phys. Chem. 76 2947 because the oxygen background in the system (1972). 2. J. A. Roberts, Jr., and A. W. Searcy, was too high. High Temp. Sci. 4_, 411 (1972). 4. SCATTERING IN A MOLECULAR BEAM AS A 1 5. GASEOUS THALLIUM (I) METABORATE AND RESULT OF INTRABEAM COLLISIONS THALLIUM (I) ALUMINUM FLUORIDE David J. Meschi Alfred Buchler and David H. Feather Collisions between molecules in a molecu While all the thallium (I) halides have lar beam cause scattering which results in been studied, only two gaseous ternary com a significant attenuation of the beam pounds of monovalent thallium, thallium (I) intensity in some cases. By the use of a nitrate and thallium (I) sulfate, have been simple model an expression can be derived studied so far. The purpose of this work for SAB, the fraction of A molecules that was to extend the range of known gaseous will traverse the beam without sustaining ternary compounds of thallium (I) and to collisions with B molecules: compare their mass spectrometric behavior to that of the corresponding alkali compounds. exp[-^TOAB2*ABCaA01/2]- "AB m The vapors above thallium (I) metaborate and a equimolar mixture of thallium (I) fluoride and aluminum fluoride were examined Here pjj is the partial pressure of B mole in an Atlas CH-4 mass SDectrometer. The cules m the source, OAB is the collision mass spectrum of thallium (I) metaborate is radius for an A-B collision, a is the area similar to that of th, alkali metaborates. of the orifice, and k and f have their Both thallium (I) metaborate monomer, TIBO2, usual meaning of the Boltzmann constant and and dimer, Ti-l(^>2)2> are present in the temperature respectively. The quantity *AB vapor, the latter identified by means of the is the value of |VA-VB|/VA averaged over the fragment ion Tl2B02+. The dimer parent ion B molecules in the beam and the A molecules Tl2(B02)2+ was not found; it was present to passing through the beam. The beam length 2 the extent of less than 1 part in 400 of the is assumed to be much greater than (a/Vp/ . fragment ion Tl2B02+. For gaseous thallium (I) nitrate dimer the fragment ion, Tl2N03+, An exact value for $AB would be extremely was also the only ion observed. In this difficult to obtain because of the non- respect the two thallium pseudohalides isctropic nature of a beam, but it can be differ from the thallium halides, which, in approximated by the expression contrast to the alkali halides and pseudo halides, contain parent diner ions in their 1/2 "AB 1/2 mass spectra. *AB » 2(1+VAB) "AB Cln>AB)wr In the vapor above the TIF - AIF3 mixture, the mixed thallium (I) aluminum fluoride, (2) T1A1F4, is present. In addition the presence of the ion Tl2AlF4+ shows the existence of 7. THE CONDITIONS OF CHEMICAL EQUILIBRIUM a higher polymer. In analogy with the assignment of the corresponding ions in the Alan W. Srarcy UF-AIF3 and NaF-AlF3 systems, we are inclined to assign this ion to the double Accepted thermodynamic theory assumes dimer Tl2(AlF4)2. In the case of T1A1F4, chemical equilibrium to rt^uire that the the parent ion TlKLTn* is present in the mass chemical potential pi of each component i spectrum, in contrast to L1AIF4 and NaAlF4, be the same throughout the system. A thermo in whose mass spectra only the fragments dynamic theory is developed in \foich a more LiAlF3+ and NaAlF3 are observed. restrictive constraint is postulated. Ecch chemical component is assumed to be character The difference in behavior between the two ized by an average molar free energy content sets of compounds nay be explained by which must have the same value in each phase reference to recent work on T1C1 and TIBr, of a system at equilibrium. Subsets of a which has shown the highest lying orbital chemical component in non-ideal phases at to be a thallium s-orbital.* Removal of equilibrium may have different sets of this s-electron on ionization will actually accessible energy states from those available strengthen the Tl-X ionic bond, the ion to other subsets; as a consequence, the formed being effectively Tl^X". Removal of average molar free energies for a component a halogen, p-electron removes the charge on i of such phases may differ from MI when the the halogen and leads to the fragment ion average molar free energies are equal. Tl+. This analysis may be extended to TIAIF4 and the alkali aluminum fluorides if The new theory is consistent with the laws we assume that the mixed thallium halide of thermodynamics, and it predicts the same has the same structure as NaAlF4, which has conditions of equilibrium as accepted theory two bridging fluoiine atoms between the for homogeneous reactions in dilute solutions, sodium and aluminum atoms. Removal of a even if imperfect, and in heterogeneous thallium s-electron leads to the parent ion equilibria of all components between perfect TIAIF4 , whereas no such process is possible solutions or of solutes between dilute in the alkali compounds. Removal of a solutions. But conditions of heterogeneous bridging fluorine p-electron, on the other equilibria involving one or more non-ideal hand, leads to the fragment ions MA1F3+ in solutions are predicted to be different from both the thallium and alkali compounds. conditions predicted by accepted theory. Inconsistencies which have been reported between values of m calculated from calori- 1. J. Berkowitz, J. Chem. Phys. 56, 2766 metric data for metallic solutions and values (1972); J. Berkowitz and J. L. Uenmer, ibid., from vapor pressure or emf measurements are 57, 3194 (1972). predicted by the new theory. 8. THE RELATIONSHIP BETWEEN FORWARD AND 6. THE ZEEMAN EFFECT IN Se2 REVERSE REACTION RATES UNDER NON-EQUILIBRIUM Alfred Biichler CONDITIONS Alan W. Searcy, Alfred Buchler and In consultation with J. T. Hougen (National Dario Beruto Bureau of Standards) a scheme for the calculation of the Zeeman effect in Se2 has One of the most important questions of been developed. The exact equations for the reaction kinetics is the circumstance under rotational energy levels for a case between which the ratio of the rate constants for a Kund's cases (b) and (c) have been derived forward and reverse reaction can be equated and will be used to calculate the unperturbed to the equilibrium constant for the reaction energy levels for Se2. If the spacing of in question. But theoretical proof of the the rotational levels is sufficiently large, relation apparently had been given only for a perturbation treatment will then be used ideal solutions. to determine the effect of the magnetic field on the energy levels. Magnetic moments It is shown that for a given elementary for the various rotational states can then reaction, and under nonequilibrium conditions, be derived by determining the variation of the ratio of the rate constant of the forward energy with applied magnetic field. reaction to the rate constant of the reverse reaction is the equilibrium constant, as long as both forward and reverse reaction take -158- place under the same experimental conditions A study of the thermodynamics of SrSCfy ans as long as products and reactants establish decomposition will be initiated. No measure an equilibrium distribution of states to the ments of this reaction have been reported. transition complex of the reaction. We hope later to measure the rate of decom position of SrSC*4 single crystals for com The proof has an immediate application parison with the thermodynamic data. Only to vaporization kinetics. As a consequence two substances, calcite (CaC03) and barite of the proof it is possible to eliminate one (BaSOaj), have been subjected to kinetic difficult- to-measure variable from an studies of the type we plan. equation that relates pressures measured in effusion studies to the equilibrium vapor A mass spectrometer study of the kinetics pressure of a material inside an effusion of decomposition of nitric oxide on tungsten chamber. The proof also establishes means and of its reaction with tungsten at high for correcting experimental rate data in temperatures has been initiated. The study solution reactions for the effect of the is of interest both because of the importance bad: reaction as equilibrium is approached. of oxides of nitrogen in air pollution and because of the insight to heterogeneous reaction kinetics that can be obtained from comparison of the results with earlier i. RESEARCH PLANS FOR CALENDAR YEAR 1973 measurements of oxygen-tungsten reactions. Alan N. Searcy If funding permits, we will initiate an attempt to measure by a differential method Theoretical work on the thermodynamics of the changes in vapor pressure of a liquid, solutions and of surfaces and interfaces probably water, produced when the liquid is will be continued. in contact with a large area of an active solid surface. According to accepted surface We plan exploratory mass spectrometer thermodynamic theory, there should be no inve tigation of the feasibility of deter difference so long as the liquid surface mination of the changes in composition and remains plane, but an experimental test has partial pressures that occur in order to apparently not been made. Funding limitations displace a phase of narrow composition range have forced abandonment of plans to study to its new composition of congruent sublima the details of kinetic energy transfer from tion .ifter a change in temperature. If the hot surfaces to cold gases. method proves satisfactory we will be able to make direct measurements of the partial molal Gibbs free energies of the phases at 10. 1972 PUBLICATIONS AND REPORTS equilibrium near their congruent compositions. Comparison of these values with values Alan W. Searcy and Associates calculated from directly measured activities of the components in the vaDor phase would Journals provide a new e>perimentai test of the thermodynamic theory of heterogeneous equili 1. Alfred Buchler and Alan W. Searcy, The brium. The studies might air.o lead to a Formation of Vapor Phase Nuclei in Nucleate convenient method for making scall changes Boiling, Mmatshefte Sir Chemie 103, S03 in compositions and therefore of electrical (1972) (article in German). English version, properties of semiconductors. UCRL-20552). We are completing a paper on the inter ?. James A. Roberts, Jr., and Alan W. Searcy, pretation of the terrace-ledge-'cink (t-l-k) The Stabilities of Ce2F6(g) and La2F6(g), model of vaporization of solids in terms of High Temp. Scl. 4_ [5], 411 (1972). transition state theory. The theoiies are shown to be complementary, but seme formula 3. David H. Feather, Alfred Buchler, and tions of mechanisms in terms of the t-l-k Alan W. Searcy, The Vapor Pressures of Gallium are inconsistent with transition stare theory. Trifluoride Monomer and Dimer, High Temp. Sci. 4_ [4], 290 (1972). We hope to complete the study of the influence of carbon dioxide pressures on 4. David J. Meschi, Scattering in a Molecu calcite decomposition that was initiated lar Beam as a Result of Intrabeam Collisions, in 1971. The study should yield data both J. Phys. Chem. 76, 2947 (1972). on the reaction kinetics and on the heat of fcrnation of a metastsble form of calcium LBL reports oxide that was identified during our calcite kinetics studies. 1. Alan W. Searcy, The Conditions of Chemical Equilibrium, LBL-898, July 1972. -159- 2. Harry B. Skinner and Alan W. Searcy, 4. Peter P. Bihuniak, The Mechanism of ZnC Mass Spectrometric Studies of Gaseous Oxides Single Crystal Vaporization, {M.S. thesis), of Rhenium, LBL-1139, October 1972. LBL-867, June 1972. 3. Roy Tom Coyle, Vaporization Kinetics of 5. David Hoover Feather, Mass Spectrometric Magnesium Nitride (Ph.D. thesis), LBL-136, Studies of Inorganic Vapors (Ph.D. thesis), January 1972. LBL-844, September 1972. 6. Pirooz Mohazzabi, The Kinetics of Vaporization of Barium Sulfate, (M.S. thesis), LBL-1416, December 1972. -160- 8. MICROSTRUCTURE AND BEHAVIOR OF CERAMIC MATERIALS: GLASS AND CERAMIC-METAL SYSTEMS Joseph A. Peak, Principal Investigator 2. J. F. MacDowell and G. H. Beall, Immiscibility and Crystallization in Al2Oj- A continuing overall objective of the SiC>2 Glasses, J. Am. Ceram. Soc. 52 [1], research program is to contribute to the 17-25 (1969). development of a fundamental understanding 3. N. L. Bowen and J. W. Greig, The System: of the factors involved in obtaining ceramic Al20,-Si02, J. Am. Ceram. Soc. 7 [4], 238-254 materials with controlled character [which and [5], 410 (1924). includes microstructure), and of the relation WEIGHT * AI O ship of character to mechanical behavior at 2 s room and high temperatures. This objective involves studies on the kinetics and mechanisms of solid state reactions which in general play a part in the development of microstructure. It also involves studies on the mechanisms responsible for the mechanical behavior of single crystals, and on the application of such knowledge to the under standing of the behavior of polycrystalline ceramic materials. A second objective of this program is con cerned with structural, thermodynamic, and electrochemical studies of glass, glass-metal, and ceramic-metal systems. It involves studies related to wetting, bonding, and the nature of the interfaces between dissimilar phases; to the thermodynamics and kinetics of chemical reactions at such interfaces; and to the kinetics and mechanisms of dis solution and diffusion in glasses. 1. STABLE AND METASTABLE EQUILIBRIA IN THE SILICA-ALUMINA SYSTEM MOLE % A1209 Ilhan A. Aksay and Joseph A. Pask Fig. Sit^-AljOj equilibrium phase dia gram as determined from the interfacial com Sapphire-fused silica diffusion couples positions of sapphire-fused silica diffusion annealed in sealed Mo crucibles in the temper couples. Mullite liquidus is by Davis and ature range of 167B to 2003°C were analyzed Pask.1 Metastable immiscibility and spinodal by electron beam microprobe. Interfacial regions are by MacDowell and Beall.2 compositions across the phase boundaries were used to determine the stable composition range (XBL 729-7039) of the mullite (3Al2 semi-infinite fused-Si02-sapphire couples WEIGHT % Al203 above 1634°C. The interfacial liquid com 40 60 80 positions provided data for a minor revision I ' I ' I ' I of the mullite liquidus curve. Diffusion ACTIVATION ENERGY, 0, FOR coefficients calculated from the Al profiles o DIFFUSION ( Abstracted from J. Am. Ceram. Soc. 55 [10], 525-531 (1972). Present address: Department of Materials Engineering, North Carolina State University, Raleigh, N.C. 27607. 3. DIFFUSION STUDIES IN THE Si02-Al2C>3 SYSTEM IN THE TEMPERATURE RANGE OF 1800 TO 2000°C o Ilhan A. Aksay and Joseph A. Pask _l_ I • l l l ' ' l • 40 60 80 Boltzmann-Matano analysis was used to MOLE %AI 0 analyze the diffusion profiles of the sapphire- 2 3 • i I , l i I i I fused silica couples annealed in the tempera O 02 04 0.6 0.8 1.0 1.2 1.4 L59 ture range of 1803 to 2003°C. The diffusivi- ALUMINUM CONCENTRATION IN MELT (gm/cm3) ties of aluminum cation in aluminum-silicate melts were concentration dependent and varied Fig. 2. The variation of activation energy by three orders of magnitude at a given with concentration for aluminun ion diffusion and viscous flow1.2 in S1O2AI2O3 melts. The diffusion data below 1870°C are by Davis and Pask.3 QffiL 729-7038] temperature (Fig. 1). The activation energy for diffusion obtained from these data decreases from 211.1 kcal/mol to ~ 40 kcal/mol with increasing alumina content of the melt (Fig. 2). An analogous relationship also DIFFUSIVITY. Dlel exists between the activation energy for TEMP.t (L^OBI Wrf C„>QBIcm/an3 viscous flow and the alumina content of the !C»>IU'eipS22[AI] 534>n*'Bp3iaiU 1 2 17a.|CT*ejp6ll m 296'ICTTeip354W melt. ' The diffusion mechanism of low i&.icr'eipesslAO \mita1t»AZ3& 341 ilO*«.p806W1 alumina contents is believed to be a coopera S42>D*>eiplio4IAl] tive movement of large oxygen-containing aluminum and silicon complexes in a type of ALUMINUM CONCENTRATION W MELT (am/em3) ring mechanism. The size of the aluminum- containing -liffusing units decreases gradually WEIGHT % Al O, z with increasing alumina content up to Fig. 1. Computer-determined diffusivity of ~ 40 mol i AI2O3. Above this concentration, aluminum ion vs concentration in aluminum- structural changes cease with the complete silicate melts in the temperature range of breakdown of the silica network. 1803 to 2003°C. (XBL 729-7040) -162- 1. R. Rossin, J. Bersan, and G. Urbain, Present address: Department of Materials Viscosity of Fused Silica and Slags Belonging Engineering, North Carolina State university, Raleigh, N.C. 27607 to the Si02-Al203 System, Compt. Rend. Acad. Sci. Paris 258 [2], 562-564 (1964]. 2. B. S. Mitui and Yu. A. Nagibin, Viscosity 1. J. J. Rasmussen and R. P. Nelson, Surface Tension and Density of Molten AI2O3, J. Am. of Al203-Si02 Melts, Izv. Vyssh. Ucheb. Zaved., Chern. Met. 12 [7], 8-10 (1969) Ceram. Soc. 5£ [8], 398-401 (1971)'. (Russian). UCRL-Trans-1479, July 1972. 2. Y. Bottinga and D. F. Weill, Densities 3. R. F. Davis and J. A. Pask, Diffusion and of Liquid Silicate Systems Calculated from Reaction Studies in the System AI2O3-S1O2, Partial Molar Volumes of Oxide Components, J. Am. Ceram. Soc. 55_ [10], 525-531 (1972). Am. J. Science 269 [2], 169-182 (1970). WEIGHT % Al20j 4. DENSITIES OF Si02-Al203 MELTS Ilhan A. Aksay, Robert F. Davis, and Joseph A. Pask An X-radiographic technique developed by Rasmussen and Nelson* was used to determine the densities of the Si02-Al203 melts in the temperature range of 1700 to 2000°C for the entire binary system. The technique con sisted of measuring the volume of a melt at temperature from its radiographed image obtained with a 300 kV X-ray source and relating this volume to its density. The densities varied linearly with tempera ture within the temperature range studied. These data were used to determine the variation of the melt densities with alumina content at a given tenperature (Fig. 1). In the temperature range of 1800 to 2200°C, the densities increase linearly up to ~ 4.r inol $ AI2O3. The negative deviation from linearity at higher alumina contents is indicative of the formation of a "looser" silicate structure. Densities were also calculated by using the partial molar volume data2 of AI2O3 and Si02 which were determined from alkali and alkaline-earth aluminum-silicate densities. The calculated densities are lower than the measured ones (Fig. 1). This discrepancy is explained by the breakdown or opening up of the silicate network when alkali and alkaline-earth oxides are added to silica. Aluminum ions, on the other hand, act as net work formers on going into four and/or six fold coordination. Thus, a higher oxygen density or smaller partial molar volume is realized in the absence of alkali and alkaline-earth metals. The presence of MOLE % Al 0 aluminum in four- and six-fold coordination 2 3 was verified by X-ray spectroscopy. Fig. 1. Densities of Si02-Al203 melts vs IIBI i AI2O3 in the temperature range of 1800 to 2200°C. The data points illustrated with Presented at the 25th Pacific Coast Regional solid triangles are calculated densities at Meeting of the Am. Ceram. 'Soc., Oct. 25, 1900°C based on the partial molar volume data 1972, Portland, Ore. of Bottinga and Weill.2 (XBL 729-7036) -163- 5. PERMEATION OF SILICATE LIQUIDS INTO prepared by calcining freeze dried powders SINTERED MAGNESIA COMPACTS formed from an aqueous solution of magnesium sulfate hydrate and aluminum sulfate hydrate. Boon Wong and Joseph A. Pask The freeze dried powder was characterized Fundamentals of permeation kinetics of a by X-ray diffraction analysis, size distribu liquid into a porous solid compact and the tion analysis, and scanning electron corresponding microstructure changes in the microscopy. Thermal reactions of the freeze solid due to the permeating liquid, either in dried powder were studied by using differen the presence or absence of reactions, were tial thermal analysis, thermogravimetric studied from both theoretical and experi analysis, and X-ray diffraction. The forma mental viewpoints. tion of spinel was postulated to be an aluminum sulfate decomposition activated Equations for capillary penetration were reaction. These postulated reactions were developed based on Poiseuille's law for verified by calculations of their free energy capillary flow and the generalized thermo change. Freeze dried powders were calcined dynamic concept of wetting of a solid by a isothermally for one hour at temperatures from liquid. A dihedral angle equation used for 900 to 1400°C. Above 1200°C, a much faster describing phase distribution (microstructure) spinel crystallite growth was observed, in a solid-liquid system was thermodynamically indicating two different growth mechanisms derived and generalized. were operating below and above 1200°C (Fig. 1). The same freeze dried powders Experimental studies were performed on the calcined at 1000 and 1300°C for 0.2 to 12 permeation of a number of silicate liquids hours showed that the spinel crystallite into sintered magnesia (MgO) compacts of size increased very slowly after 2 hours of about 92.5% theoretical density. The experi calcination. mental results support the theoretical predictions. The reactivity of the freeze dried powder to form spinel on calcination depended upon Permeation rate of a viscous liquid into the freeze drying temperature and the starting a "homogeneous" compact with fine capillaries solution concentration. Freeze dried powder depends mainly on the driving force for prepared from a starting solution containing wetting, which in turn depends markedly on the presence and rate of any interfacial IOOO chemical reactions. In addition, the average permeation distance of the viscous liquid into the compact increases parabolically with time during the initial stages. A negative deviation from this parabolic relationship usually results at later stages when the back-pressure developed by compres sion of entrapped gases becomes significant. The microstructure of a relatively dense solid compact is markedly affected by a permeating liquid. The degree of material redistribution in the solid by a solution- precipitation process is significantly dependent on the degree of reactivity of the liquid with the solid. Presented at the 25th Pacific Coast Regional Meeting of the Am. Ceram. Soc, Oct. 25, 1972, Portland, Oregon. Abstracted from LBL-877, August 1972. 6. POLYCRYSTALLINE SPINEL FROM POWERS PREPARED BY FREEZE DRYING TECHNIQUE 800 1000 IZOO Chi-Shine Tao and Joseph A. Pask CALCINATION TEMPERATURE, °C Fig. 1. Spinel crystallite size vs. calcina Stoichiometric spinel (MgAl204) powder was tion temperature. (XBL 7210-7060) -164- 12% by weight of anhydrous magnesium and aluminum sulfates and dried at 25°C was Partly presented at the 25th Pacific Coast found to be the most satisfactory. Regional Meeting of the Am. Ceram. Soc., Oct. 25, 1972, Portland, Oregon. Spinel powders developed from the freeze drying technique consisted of submicron size powders and large aggregates of 5 to 50 microns in size. Large aggregates among the 8. IMPACT FRACTURE OF BRITTLE TRANSPARENT fine powders form regions of lower green MATERIALS density resulting in the formation of a nonhomogeneous sintered microstructure. The William E. Snowden and Joseph A. Pask factors leading to the formation of large aggregates are being investigated. An investigation of the interaction of stress waves with cracks in transparent materials, primarily glass, has continued. An experimental setup consisting of a lens Presented at the 25th Pacific Coast Regional Schlieren system, high speed cameras, a Meeting of the Am. Ceram. Soc, Oct. 25, 1972, spark-discharge light source, and a magnetic Portland, Oregon. Partially abstracted from hammer for producing impacts has been LBL-876. investigated for suitability in studying the problem. Several problems related to synchronization of all phases of the event have been solved; in particular, the internal 7. LIQUID PHASE SIOTERING STUDIES* light system of the Beckman and Whitley Model 222 framing camera has been replaced Carl E. Hoge and Joseph A. Pask by a small laser, permitting timing pulses for each frame to be recorded. Equations have been derived describing kinetic changes during densification of a Flyer straps of several metals were made sintering powder compact in the presence of a for use with the magnetic hammer. Aluminum liquid phase. The models used correspond to was found to be most suitable, and varying uniform sized spherical particles forming the strap thicknesses are being used to give solid phase in simple cubic packing. The changes in the duration of the impact from liquid phase is assumed to wet the solid, and approximately 100 nsec to 1 usee. An exten some solid solubility occurs in the liquid. sive investigation of flyer characteristics Cases treated include: has been undertaken. Streaking camera records are being used to give the velocity-time- 1. Small amounts of liquid phase with displacement history as a function of zero dihedral angle. capacitor bank voltage. The data are useful 2. Large amounts of liquid phase with for determining the experimental conditions zero dihedral angle and no gas necessary to produce an impact of kutown entrapment. amplitude and duration occurring at some 3. Large amounts of liquid phase with reproducible time. zero dihedral angle and entrapped gases. 4. Small amounts of liquid with nonzero dihedral angle. A critical review of several theoretical approaches to the problem of crack stability in brittle materials subjected to impact Zero dihedral angle cases correspond to indicates that none of these adequately sintering by solution-precipitation. With describes the phenomena. It is easily shown, systems having nonzero dihedral angles, the however, that fracture occurring as a result sintering mechanism is solid state, having of stress waves must depend on the existence enhanced vacancy concentration gradients in of both a critical pulse amplitude and a the solid due to the presence of the liquid. critical pulse duration. The relation A thermodynamic treatment to determine between pulse characteristics and crack possible end-point densities in compacts length is being determined. indicates that as the dihedral angle decreases, critical end-point densities exist. Mechanisms of grain growth are also being investigated by using thermodynamic arguments. 9. MECHANICAL BEHAVIOR OF MgO SINGLE CRYSTALS Kinetic data is being obtained by using a dilatometric technique. Powder compacts, Chung Dokko and Joseph A. Pask closely approximating the parameters of the models proposed, are employed to compare The effort toward studying the alloying densification kinetics with predicted results. effect on the mechanical behavior of MgO was -165- discontinued because of the difficulty ture in the specimens of Type HP was found encountered in obtaining the desired single- to increase the yield stress and limit the crystal specimens. Attention has been total strain at fracture, presumably by directed to the high temperature deformation restricting the initiation and propagation of MgO single crystals. Specimens may be of dislocations. In the specimens of Type deformed under increasing or constant stress HPA, lower yield stresses in agreement with levels which give rise to stress-strain and values for slip in single crystals and greater creep curves, respectively. It is known that plastic strains were observed, indicating substantially different stress-strain behavior that dislocation mobility in polycrystalline is exhibited, depending on the crystal orien MgO is significantly enhanced by annealing tation and hence the operative slip planes. out defects along grain boundaries. Also, recent work on LiF single crystals of <100 > and <111 > orientation has revealed Creep results for tests conducted at 1200°C some differences in creep behavior in terms were in accord with the stress-strain of primary region, stress dependence of observations noted above. For specimens of steady-state creep rate, and substructure.1 both types, the experimental values for strain rate 6 as a function of stress a were In the present study, an attempt is being compared with predictions arising from a made to correlate stress-strain and creep number of theoretical models. Both experi behavior in MgO single crystals with different mental data and theoretical predictions for orientations tested over the temperature specimens of Type HP are presented in Fig. 1. range of 0.4-0.6 T . Stress-strain curves Creep in this material was found to be con will be obtained bym using both constant stress trolled by a grain boundary sliding mechanism; and constant strain rate conditions. Factors the predicted and experimental values for e of interest are the observed maximum strain differ by less than an order of magnitude, hardening rate at 1000°C, and the contribu and the measured stress exponent n = 1.8 is tion of (partial) recovery to the stress- in good agreement with the predicted value strain behavior. Also, preliminary investiga tions have repeatedly shown the presence of nonpropagating cracks in the deformed specimen. The effect of such occurrences together with buckling on stress-strain curves will be examined. In the analysis of creep, points of interest are whether measurable creep occurs in a single crystal at a stress level below the yield stress, and the reason for the long primary region for < 100 > crystals com pared with <111 ) crystals. 1. D. R. Cropper aid Joseph A. Pask, Creep of Lithium Fluoride Single Crystals at Elevated Temperatures, LBL-1152, October 1972. 10. EFFECT OF GRAIN BOUNDARY CHARACTER ON HIGH TEMPERATURE MECHANICAL BEHAVIOR OF POLYCRYSTALLINE MgO William E. Snowden and Joseph A. Pask The high temperature mechanical behavior of two types of polycrystalline MgO was studied in compression over the temperature 1—i i i r1111 i. i i • • 11 ••! range 1200-1400°C. Differences in behavior 4 3 2 between hot-pressed specimens (Type HP) and 10" IO" K)- specimens which were hot-pressed and annealed (0-/G) (Type HPA) were attributed to differences in Fig. 1. Experimental data and theoretical the structures of the grain boundary regions. predictions for hot-pressed specimens (Type The presence of a grain boundary defect struc HP). (XBL 727-6602) -166- 1 i i i i M i| i ii 1 1 III ~110 kcal/mol was observed. These values are in agreement with previously reported values for the activation energies for extrinsic and intrinsic diffusion of oxygen, I0"6 i II I 1 respectively. n-7l i 11. CONTROL OF GRAIN BOUNDARY CHARACTER IN POLYCRYSTAIXINE MgO S> |n«7.3 - Truett Sweeting and Joseph A. Pask 7 The nature of the grain boundary has been icr - found to be significant in controlling the /' mechanical behavior of polycrystalline MgO.1-3 Factors that determine the grain ~ P ' ' - - / ' ' - boundary structure in processing are being / ; ln'4.5 studied. Resulting microstructures are being - analyzed with emphasis on grain boundary nature. Stress-strain tests will be made to 1 $1 relate the resulting structures to mechanical behavior. 8 io- 1 • 1. S. M. Copley and J. A. Pask, Deformation of Polycrystalline MgO at Elevated Tempera tures, J. Am. Ceram. Soc. 48 [12], 636-642 I? (1965). i 1 1 1 111 l/l 3 1 II 2. T. G. Langdon and' J. A. Pask, Effect of lO"" IO" \C Microstructure on Deformation of Polycrystal (0-/G) line MgO, J. Am. Ceram. Soc. 54 [5], 240-246 Fig. 2. Experimental data and theoretical (1971). predictions for hot-pressed and annealed 3. W. E. Snowden, High Temperature Mechani specimens (Type HPA). (XBL 727-6601) cal Behavior of Mg0-Ca0-Mg0-Si02 Materials (M.S. thesis), UCRL-20567, July 1971. n = 2. For the specimens of Type HPA, experi mental data and predictions arising from the Weertman dislocation glide and climb model 12. INTERFACIAL REACTIONS AND WETTING are presented in Fig. 2, Creep in this BEHAVIOR OF GLASS-IRON SYSTEMS* material was found to be controlled by a dislocation climb mechanism, and a stress Carl E. Hoge, John Brennan,''' and Joseph A. exponent n = 7 was measured. While this Pask value is higher than n = 4.5 predicted by the Weertman model, this dependency is sensi Interfacial reactions and enhanced wetting tive to dislocation density and breaks down or spreading of Na2FexSi205+x and at some critical stress o*. For the speci Na2-2xFexSi2°5 glasses on substrate of high mens of Type HPA, a* was found to be approxi purity iron (Marz) and commercial iron mately equal to the lowest stress (10,000 psi) (Armco) were observed at 1000°C at low used in this work. Thus the value n = 7 partial pressures of Oj and Na vapors. determined experimentally is found to be con Reactivity increased with increase of the sistent with the Weertman model in the regime oxygen/silicon ratio in the glass and purity where an exponential relation between E and of iron, and with decrease of the gas pres- a is predicted. The apparent activation eneTgy Q for creep was determined by using an incremental temper ature technique. For specimens of Type HP, Abstracted from LBL-480, July 1972. a value for Q of ~ 70 kcal/mol was found, Present address: United Aircraft Research while for speciiaens of Type HPA a value of Labs., East Hartford, Conn. 06108. -167- 13. EFFECT OF COMPOSITIONS ON GLASS-METAL Studies will also be undertaken on the INTERFACE REACTIONS AND ADHERENCE* mechanisms and kinetics of interdiffusion and moving boundaries in single crystal oxide John Brennan+ and Joseph A. Pask couples. Diffusion data will be interpreted in terms of thermodynamic properties of the Reduction-oxidation reactions and enhanced system. wetting or spreading of Na2FexSi20s+.x, Na2CoxSi205+x, and N^NIxS^Os+x glasses on substrates of Fe, Co, Ni, NiFe, and NiCo were b. Development of Microstructures observed at 1000°C at low partial pressures of O2 and Na vapors as the O/Si ratio of the Studies of the factors that play a role glass was increased. When the substrate in liquid phase sintering and in the develop metal had a higher oxidation potential than ment of microstructure are continuing. The the metal of one of the cations in the glass, approach will consist of selecting systems e.g., CoO containing glass on Fe, redox in which a range of dihedral angles will reactions resulted in the formation of exist. Thermodynamic, as well as kinetic, metallic precipitates under all conditions. arguments will be applied. A third type of redox reaction based on reduction of the valence of a cation in the Freeze drying techniques are continuing to glass, e.g., Fe3+ to Fe2+, also occurred. be utilized in the preparation of powders Adherence developed between substrates and for purposes of attaining polycrystalline glasses containing amounts of substrate oxide materials with controlled composition and approaching saturation either in the starting microstructure. Particular attention will composition or attained by redox reactions. be paid to the optimization of the process to prevent the formation of aggregates, or to break them up easily, in order to realize a completely uniform texture in the compacted Abstracted from LBL-481, July 1972. powders. Present address: United Aircraft Research Labs., East Hartford, Conn. 06108. c. Mechanical Behavior An investigation of crack propagation in brittle transparent materials subjected to 14. ELECTROCHEMICAL STUDIES OF THE high velocity impacts is continuing. The Fen.95O-Na20-SiO2 GLASS SYSTEM phenomenon of branching in a brittle material subjected to transient stresses will also Robert B. Langston and Joseph A. Pask be investigated. Glass specimens of several ternary com Stress-strain and creep data will be ob positions have been prepared in preparation tained at high temperatures on MgO single for galvanic cell measurements at tempera crystals. These will be correlated for tures of 700 to 1030°C. These have been purposes of evaluating deformation mechanisms analyzed chemically by using atomic absorption and the conditions under which internal techniques. These eX[.-eriments will also cracks form. provide information on the liquidus surface of the ternary phase diagram Fen gsO-Na20- The stress-strain behavior of polycrystal Si0 . 2 line MgO specimens with different grain boundary characteristics will be evaluated. Specimens will be prepared with the nature of the grain boundary as a variable. Some 15. RESEARCH PLANS FOR CALENDAR YEAR 1973 stress-strain data will also be obtained on polycrystalline spinel. Joseph A. Pask a. Solid State Reactions d. Glass Studies Diffusion experiments designed to determine Determinations of activities of iron oxides the chemical interdiffusion kinetics of in sodium silicate glasses hy electrochemical aluminum and silicon ions in mullite are methods, and calculations of thermodynamic continuing. The effect of solid solution properties are being continued. Associated additives to mullite on diffusion kinetics information will assist in checking the will also be determined. FeO-Na20-Si02 phase diagram. -168- 16. 15"2 PUBLICATIONS AND REPORTS LBL reports Joseph A. Pask and Associates 1. C. E. Hoge, J. Brennan, and J. A. Pask, Interfacial Reactions and Wetting Behavior Journals and books of Glass-Iron Systems, LBL-480, July 1972. 1. Robert F. Davis, Ilhan A. Aksay, and 2. J. Brennan and J. A. Pask, Effect of Joseph A. Pask, Decomposition of Mullite, Compositions on Glass-Metal Interface J. Am. Ceram. Soc. 5S [2], 98-101 (1972) Reactions and Adherence, LBL-481, July 1972. (UCRL-20536). — 3. C. C. Tao, Polycrystalline Spinel from 2. Joseph A. Pask, Chemical Reactions and Powders Prepared by Freeze Drying Technique Adherence at Glass-Metal Interfaces, in (M.S. thesis), LBL-876, June 1972. Proceedings of the Porcelain Enamel Institute [Technical Forum 1971}, 33, 1-16 (1972) 4. B. Wong, Permeation of Silicate Liquids (LBL-166). — into Sintered Magnesia Compacts (M.S. thesis), LBL-877, August 1972. 3. Robert F. Davis and Joseph A. Pask, Diffusion and Reaction Studies in the System 3. W. E. Snowden and J. A. Pask, Effect of Al;>03-Si02, J. Am. Ceram. Soc. 55 [10], Grain Boundary Character on High Temperature 52S-S31 (1972) (UCRL-19644 Rev."T). Mechanical Behavior of Polycrystalline MgO, LBL-1120, December 1972. Invited papers presented 6. D. R. Cropper and J. A. Pask, Creep of 1. 'ttfetal-Ceramic Bonding," A.I.M.E. No. Lithium Fluoride Single Crystals at Elevated Calif. Metallurgical Section Symposium on Tejperatures, LBL-1152, September 1972. Joining of Materials, Palo Alto, March 18, 1972. 7. Abbaraju P. Raju, Ilhan A. Aksay, and Joseph A. Pask, Permeation of Silicates in 2. Chair.ran, U.S.-Japan Seminar on Basic Magnesia and Forstsrite Conpacts, LBL-159 Rev., Science of Ceramics, "Equilibria and Kinetics May 1972. in Modern Ceramic Processing," Berkeley, April 28; Niagara Falls, N.Y., May 1; Penn State Univ., May 3-5, 1972. 3. "Nature of Glass-Metal Interfaces," Symposium on Interfacial Bonding and Fracture in Polymeric, Metallic and Ceramic Composites, U.C.L.A., Los Angeles, Calif., Nov. 13-15, 197?. -169- C. RELATION OF MICROSTRUCTURE TO PROPERTIES IN CERAMICS Riahard M. Fulrath, Principal Investigator 1. HOT STAGE SB! STUDIES OF THE INITIAL STAGE SINTERING OF COPPER AND NICKEL Craig B. Shumaker+ In the past, experimental procedures for studying isothermal sintering have used dilatometers to measure shrinkage of compacts, electron images to measure neck growth and the length change of lines of spheres, and the hot stage optical microscope to measure neck growth and two-sphere shrinkage. The recent developments in scanning electron microscopy by Fulrathl have allowed continuous observa tion of the sintering phenomena at tempera tures up to 1600°C with magnifications from 50 to 5000*. Initial work on the first stage sintering of copper and nickel microsphere compacts using data obtained from the SEM is completed. Shrinkage measurements were made from photo graphs of the compact surface taken from the SEM television monitor at low magnifica tion. Isothermal sintering experiments have shown that considerable erratic linear change occurs before continual shrinkage commences. Figure 1 shows typical sintering behavior of a -30+20 urn nickel sphere compact of 434 green density. The AL/Lo calculations, linear least square slopes, graphs and time-length corrected data calculations were made by computer analysis. to 800 .TOO sno K)oo Tt.'.ii-ERATURE (t) After time-length corrections to eliminate 0 20 5D 70 90 TIME (MINI 1 I 1 1 1 Fig. 2. Dimensional changes in heating nickel compacts under identical conditions. Ni-A Ni-1 was compacted to a green density of 3.0 421 TD and Ni-2 was SOI TD. Solid line is r^^ the change expected from thermal expansion with no shrinkage and dotted line is a least square slope of the experimental data. 1.0. - CXBL 727-6669) erratic initial sintering the shrinkage rate 0.3 Cslope) obtained from eight isothermal sintering experiments do not agree with exist ing sintering theories. Previous shrinkage t theories predict AL/Lo proportional to ' ' Ctime)"-31"u-50 depending on diffusion paths to the neck surface. This study placed the time exponent (slope) between 0.60 and 0.70. Fig. 1. Isothermal sintering of a compact of Ni microspheres (slope = 0.609). By superimposing photographs of t = 0 and CXBL 727-6664) t = 200 minutes permanent particle shifts on -170- the surface of the compact were observed. Theories to Real Systems," presented at the By replotting the shrinkage data without the Third International Conference- on Sintering, measurements that used locations that shifted Notre Dame University, Indiana, June 1972. permanently, the initial erratic sintering and the need for 3 time-length correction were avoided. Preselection of data only decreased the standard deviation of the slope; the shrint-age rate itself remained essentially 2. SINTERING CERAMIC OXIDES unchanged. However, the error in the average AL/Lo increased at initial times, indicating David N. K. Wang and Richard M. Fulrath that even so-called good measuring locations had shifted during the first few minutes of Submicron aluminum oxide particles com sintering. pacted to nearly 40? TD (theoretical density) have been heated in the hot stage of the The observation of particle rearrangement scanning electron microscope. The associated supports the theory of Exner et al.2 that sintering, both during heating and while density gradients in real compacts cause non being held at temperature, was determined by uniform sintering. Regularly packed volumes the use of markers (tungsten microspheres, are connected by bridges of spheres that have +10-20 urn diameter) placed on the compact necks of unequal radii of curvature. Mass surface. transport is enhanced on the side of the neck of the smaller radius of curvature The specimens were supplied by the General causing the bridge (or chain) of spheres to Electric Company, Cleveland, Ohio and con straighten. Hence, uniform shrinkage that sisted of Linde "A" alumina (0.3 um average occurs in the close-packed regions can be particle diameter) doped with 0.1 wt % MgO undetected by chain straightening of spheres compacted into disks and prefired to 900°C in adjacent open regions of the compact. It for one hour to remove organic pressing aids. is postulated that sphere to sphere shrinkage Specimens for the scanning electron micro theories cannot be applied to real compacts scope study were small disks 0.25 in. without considering the influence of the diameter and 0.040 in. thick cut from the initial green density of the compact. original prefired disks. In a preliminary attempt to characterize The heating cycle followed was identical the effect of green density or initial stage for the two specimens. The samples were sintering of compacts, two experiments were first heated to 800°C and held one hour. run using nickel microspheres +20-3(1 ym with They were then heated at 4.2°C/min to 1560"C green densities of 42 and 50%. The 42% green to ensure thermal equilibrium in the hot dense sample showed nearly no shrinkage stage. The samples were then held at 1560°C despite being at adequate sintering tempera for 100 minutes. ture (Fig. 2a), whereas Fig. 2h shows con siderable shrinkage for the 50% dense sample Figure 1 shows the average linear shrinkage using identical experimental conditions. based on 10 measurements at various times in Therefore, it is concluded qualitatively the heating cycle for two specimens. As that increased green density reduces the expected, when 900°C is selected as the extent of localized sintering. initial dimension the compact expands due to thermal expansion until the onset of sintering, which occurs at approximately 1170°C. The shrinkage rate increases until a maximum Abstracted from C. B Shumaker, Initial rate is reached at approximately 1450°C. On Stage Silvering of Copper and Nickel. (M.S. thesis), LBL-1110, Aug. 1972. further heating snd during the isothermal f period the rate continously decreases. The Present address: Purdue University, total shrinkage observed and room temperature Lafayette, Indiana measurements on the specimen after the sintering run indicate a density increase 1. R. M. Fulrath, Scanning Electron Micro from approximately 40 to 90% TD. scopy to 1600°C, in Scanning "lectron Microscopy/1972 (Part I), III' Research The data obtained to date on sintering Institute, Chicago, April 1972. of oxides indicate that the scanning electron microscr • will be a powerful tool in 2. H. E. Exner, G. Petzow, and P. Wellner, studying . itering throughout the whole "Problens in the Extension of Sintering process. -171- A Meon value of 10 measurements (110 x) Run I O Mean value of 10 measurements (60x) Runs »-&r Pbre coalescence and - elimination with concurrent grain growth S5 10 Tf "2 ^ 14 < ,6 18 Constont heating rote Constant temperature Temperature <*C) |m'0 COO 1100 1200 1300 1400 I50OI56O 50 100150 200 Time (minutes) Fig. 1. Shrinkage characteristics of sub- micron particle AI2O3 compacts as determined by hot-stage scanning electron microscopy. Both runs show identical results. (XBL 731-2049) 3. INTRINSIC AND EXTRINSIC NONSTOICHICMErRY In the present study the range of the IN THE LEAD ZIRCONATE-TITA.WE SYSTEM* intrinsic and the nature of the extrinsic nonstoichiometry in lead zirconate-titanate Robert L. Holman'*' has been characterized. In addition, a simple method has been developed to allow precise Polycrystalline lead zirconate-titanate and reproducible control of the defect (PZT) ceramics are of commercial interest structure. because of their wide range of dielectric constants, pronounced ferroelectric, Two different gravimetric experiments have piezoelectric, and electro-optic behavior, led to similar values for the Pb (TixZri-x)03 coupled with low cost of processing and intrinsic single-phase region width. manufacture. Several investigators have shown that significant changes in physical properties a. Vapor Phase Equilibration Experiment can result from only slight compositional changes, or deviations in stoichiometry A single-phase, high purity, and porous created during processing by the evaporation PZT sample was suspended within a constant cf lead oxide (PbO). This causes difficulties activity multi-phase (CAMP) crucible at in the manufacturing of reproducible PZT constant temperature (1100°C) by a wire from materials ef controlled composition. a continuously recording microbalance. Weight changes are observed as a consequence Jaffe et al.l sought to inhibit this of the reaction of the sample with the CAMP undesirable evaporation of PbO by sintering crucible atmosphere. When the sample's PZT samples in the presence of the PbO weight becomes constant, the crucible and atmosphere provided by lead-oxide-enriched the sample will be in a state of quasi- lead zirconate discs. Dungen et al.6 used equilibrium. c burial powder of the exact composition as the sample to produce the necessary local As an example, the proposed PbO-TlO? phase equilibrium environment. diagram and its corresponding lead oxide isothei.nal (1100°C) activity iHagram is Although recent experimenters have been shown in Fig. 1, assuming a small degree of able to obtain improved results using these nonstoichiometry. Composition A, high approaches, stoichiometry variations and PbO CAMP, refers to a two-phase mixture of inhomogeneities were still observed. PZT with excess PbO. Composition B, low -172- crucible, assisted by the high vapor pressure 1800 of PbO at temperature, causes the sample to 1600 gain weight (PbO) until its activity is raised to equal that within the crucible of 1*30 PbO activity A. Weight gain will stop at composition point D, fixing one side of the 1100 single-phase region. The sample will con ^IjQUID+PtTiOa | tinue to maintain this weight for as long as 800 the atmosphere within the crucible remains 600- psnOj+TiOi.,, at a constant PbO activity. 'ro0 - PUOsst POTi0 3 Next, this sample, now with PbO activity 200- D, is equilibrated with a low CAMP crucible _L_ _l L_ -I l_ _l_ of composition and PbO activity B. This new P60 10 20 30 40 50 60 70 80 90 Ti0 2 PbO activity difference causes the sample to MOLE % TiCfe lose weight to the atmosphere, until point E (0) is attained. This point defines the low PbO activity side of the single-phase region. 10 Hence, the weight change (loss), D to E, ?\LtQUID PI LKJUlD+PbTlOs Pb710 *-T102„ j should exactly correspond to the molar width s of the single-phase region. A D By replacing the sample at activity E, in •-aw I the first crucible of high activity A, reversibility may be demonstrated by a weight "HIGH gain exactly corresponding to the previous 5? CAMP" weight loss. _ \ 3 oo' = "LOW b. Khudsen Effusion Experiment E B t 3 aE=oB: CAMP" The "vapor phase equilibration" (VPE) experiment measured the width of mast PZT single-phase regions while also affording a uniqu2 nethod of fixing the stoichiometry 1.0 0.5 x« 0 during processing. However, the technique MOLE FRACTION PtlO tXpto) could not be applied as easily to the solid (b) solution oppositions at/or quite near pure lead titanate or lead zirconate because of Fig. 1. (a) Proposed PbO-Ti02 phase diagram difficulties encountered in fabricating after Moon, and assuming a lead titanate mechanically stable atmosphere crucibles and region of nonstoichiometry. samples. Hence, another approach was taken. (b) Activity of PbO vs. mole fraction PbO at C constant temperature (1100 C) for the lead A variation of the standard Knudsen titanate binary. effusion cell allows for an extended Points A and B designate the constant PbO interpretation of the vapor pressure data. atmospheres provided by the CAMP crucibles In fact, a single experiment may be designed of these compositions; point C locates the to yield the width of any PZT single-phase composition and activity of the nearly region, the equilibrium lead oxide vapor stoichiometric single-phase compound, whereas pressure data as a function of temperature D and E indicate the single-phase boundaries. and composition, and the exact location of CXBL 726-6447) the stoichiometric composition, Pb(TixZri-x)03, within its single-phase region. This is possible because only a single component, PbO CAMP, refers to a multi-phase mixture of PbO, evaporates, so that the composition of PZT with excess titania and/or Urconia. the remaining condensed phase changes Conposition C designates the initial uniformly, and can be so constrained to cross stoichiometry of a single-phase •sample. The the desired phase boundaries at constant exact location of C.is dependant upon the temperature. materials's processing history. A typical Knudsen effusion experiment was The sanple, activity C, is first equili conducted where the sample is sealed within brated with the atmosphere provided by a high a non-reacting cell containing a small CAMP crucible of composition A. The activity cylindrical orifice, and the cell is con difference between the sample and tha tinuously weighed, in vacuum, as a function -173- of temperature. At constant temperature, TIME, HOURS the equilibrium vapor pressure builds up LINEAR WEIGHT LOSS REGION within the cell, and the orifice area acts [PZT <• --iauio] as the effective area from which the vapor molecules will escape at the equilibrium rate. This equilibrium escape rate is inferred from the weight-change of the cell by means of the Khudsen equation NOW- LINEAR WEI8HT LOSS REGION 1 2 [-PZT-] p = Cdw/dt)[2kT/m] '' /A0W where P = equilibrium vapor pressure dw/dt •-' the weight-loss-rate of the cell = the temperature of the orifice T LINEAR WEIGHT LOSS in degrees Kelvin REGION = mass of the effusing specie [PZT+ZtT] (PbO) Ao = the orifice area W = the Clausing correction factor for the cylinderical orifice = Boltzman's constant. TEMPERATURE CONSTANT Fig. 2. General form of the Khudsen effusion The experimental variation instituted in data at a constant temperature when the this study consisted of initially preparing Khudsen sample alters its composition through a well characterized lead zirconate-titanate the PbTixZri-^ single-phase region. This plus lead oxide powder sample, continuing the corresponds to the nonlinear portion of the Khudsen effusion experiment until all the weight-loss vs. time curve. (XBL 726-644S) lead oxide is exhausted from the cell, and then recording the weight and analyzing TIME, HOURS the oxide residue. 1 "~T"*— i-»*„ ' ! ' If the observed rate of weight loss of the V—i ' ' ' cell is constant, so is the PbO equilibrium - vapor pressure, and the composition of the sample lies in a two-phase region. When the 1- « -D ' \ ft' n1*" rate of weight loss is not constant, the " 1 * \ • "wo \o equilibrium PbO vapor pressure is changing with time and the sample contains only a .« JL._J,~,^ ^_: • single phase. Throughout a two-phase region, i the overall sample composition is changing, but the vapor pressure of PbO is not. J CONSTANT TEMPERATURE , | \ ' 1 > • ' When during the course of the experiment the lead oxide effusion causes the sample TYPICAL WEIGHT LOSS DATA FROM XNUDSEN EFFUSION EXPERIMENT OH (Pt)(Ti,Zr ,)0 + 8PllO) composition to cross into the PZT single- w 3 phase region, the weight-loss rate observed became nonlinear and slower, reflecting the Fig. 3. Khudsen cell weight loss vs. time reduced lead oxide activity in th. sample as during a typical experiment. Equilibration it becomes more deficient with respect to at successive temperatures on both sides of lead. When the continuous effusion of lead the single-phase region allows efficient oxide causes the appearance of a new second collection of vapor pressure data. AW, phase the weight-loss rate, and therefore the corresponding to the weight-loss from B to vapor pressure, will become constant once C on the activity diagram, measures the width again. Thus, the amount of lead oxide lost of the PZT single-phase region. during the nonlinear portion of the weight- (XBL 723-6075A) loss data will correspond to the exact width of the PZT single-phase region with respect Table I where "mol. and wt.4" refers to the to PbO at that temperature, or the degree percent lost. They agree closely with the of nonstoichiometry of the compound, as is data obtained by 'Vapor phase equilibration" illustrated in general in Fig. 2. techniques. It is seen that both the terminal compositions, lead titanate and lead Figure 3 summarizes the entire form of the zirconate, possess the widest ranges of non data. The actual results are summarized in stoichiometry in the system. -174- Table I. Tabulated results of the "Vapor phase equilibration (VPE)" experiment and the Knudsen Effusion (KE) experiment defining the extent of nonstoichiometry on width of the PZT single-phase region. Pb -x°x(Tis Z'l-y)°3- x^x flW SINSLE-PHASE WIDTH SAMPLE COMPOSITION (BOUNDARY) EXPERIMENTAL METHOD y WT. % "MOL.% PbO" X MOL.% PbO MOL.% D,+ MOL.% TcO? MOL.% ZrOp MOL. FRACTION PbO VPE. K.E. 1.0 7.36 10.011.0 0.100 45.0 5.0 50.0 0 0.474 X 0.9 4.01 5.5+0.5 0.055 47 25 2.75 45.0 5.0 0.486 X X 0.8 3.1 4.2+0.5 0.042 47.9 2.1 40.0 10 0.490 X X 0.65 1.89 2.7+ 0.3 0.027 48.6 j 1.35 32.5 17.5 0.493 X X 0.5 1.72 2.48+0.3 0.025 48.75 1.25 25.0 25.0 0.494 X X 0.4 1.08 1.6+0.2 0.016 49.2 0.8 20.0 30.0 ~495 X X 0.2 1.98 3.0+0.5 0.030 48.5 1.5 10.0 40.0 4925 X X 0 6.1 9.5i 1.0 0.095 45.475 4.525 0 50.0 0.475 X FORMULAE USED * (I) "MOL % PbO" s NO. MOLES MJgL (l00) = i||jMd2I> (100) 8 W = SAMPLE WT. LOSS NO. MOLES PbO IN SAMPLE W = INIT; . SAMPLE WT (GRAMS PbO LOST) (8WHI00) s (2) "WT. %PbO' (GRAMS PbO IN SAMPLE) [(W5)/(MW-PZT)] (223.19) MW-PZT= MOLECULAR WT. "WT. % PbO" = "(ML. % PbO" (4) MCL. % !>bO = ^Y* (IOC ) MOL. % TiO-* " (100) M0L.%O,* - -|-(I00) MOL.% ZrO - ^000 (5) MOLE FR";TION PbO = i-x = i-x (l-X)rl 2-X h "Mol. or wt % PbO" refers to the percent lost from sample during experiment. Assuming the rate of weight loss to be of mixings dependence on composition is shown proportional to the vapor pressure of PbO for both materials in Fig. 5. within the cell at all times, tangents con structed to the weight-loss curve allow cal In addition, each experiment generates culation of the lend oxide vapor pressure lead oxide vapor pressure data for the multi or activity as a function of composition. phase regions that bound the nonstoichioiEetric compound. The results that are obtained These data, allow the graphical integration compare favorably with those reported by of the Gibbs-Duhem equation, v-hich yields Hardtl and Rau3 for the (PZT+Z) phase field. the corresponding activity of titania or zirconia and consequently the free energy Because PbO is the only significant vapor of mixing as a Function of composition species in the PbO-Ti02~Zr02 system, it is through the PZT single-phase region. This expected that a second law analysis of the calculated free energy of mixing is found Knudsen effusion vapor pressure data should to have a minimum within the single-phase yield equivalent values for the entropies region, implying that the stoichiometric of vaporization. This is observed for PbO, composition is not the most stable perovskite. PbTi03+Ti02, and PbZr03+Zr02. This analysis was applied to both lead titanate ana lead zirconate. The reduced By plotting the lead oxide activity of experimental results for activity as a PZT in equilibrium with lead-rich liquid as function of composition are shown in Fig. 4 a function of temperature and the for PbTiOj, The result of the free energy Pb(Tix2ri-x)03 composition, it is seen that -175- 3+ + 1 i > 1 i tions of crystal chemistry (Bi and La^ 1.0 - substitution for Pb2+, Nb5+ and Sc3+ substitu tion for (Ti4+, Zr4+). T = 1091 °C Small additions of other oxides to PZT - I will alter its defect structure and ha.!d been reported to affect its electrical Oreo properties, sintering kinetics, and grain — ' J growth. Bismuth and niobium have been shown to act as sintering aids, and to improve the - ferroelectric behavior. Lanthanum has markedly enhanced the sinterability of PZT ceramics. Lee4 has demonstrated that ACTIVIT Y a— G scandium can convert the ferroelectric Tl02 i« i characteristics of PZT to an anti- 0.525 ferroelectric-like behavior. - To generate commercially useful devices, reproducibility of results must be assured, which implies that the processing kinetics 0.01 must be controlled. It appears that to ~ T:T".'.A.TE SINGLE PHASE REGION •>> explain the reported results correctly, one ^ must understand the kinetics of the incorpora , 1 . 1 tion of the Bi3+, Nb5+, La3+, and Sc3+ ions 0.50 0.51 0.52 053 i MOLE FRACTION TITANIA (X , ) into the perovskite lattice, their location T 0j in the lattice, and the defect structure or STOICHIOMETRIC LEAD DEFECTIVE ». COMPOSITION extrinsic nonstoichiometry that is generated. Consequently, an experiment was developed Fig. i. Lead oxide and titania activity as a that provided direct observation of the function of composition in the lead titanate mechanisms. single-phase region as determined from analysis of the Khudsen effusion data and a graphical integration uf the Gibbs-Duhem equation. (XBL 726-6431) approximation of PbO activity of PZT plus lead-rich liquid compositions by assuming the applicability of Raoult's law for the liquid phase is invalid. The actual activity coef ficients as calculated from the present experimental data, deviate substantially from Raoult's law. (at 1100°C YphO = -2S for PbCTi.5Zr.5)03) Finally, by evaporation of all the lead oxide from the Khudsen cell, and by carefully weighing arid analysing the sample residue that is left in the cell, it was possible to locate the stoichiometric PZT composition relative to tiie boundaries of its single- phase region. By modifying the vapor phase equilibration technique, a third gravimetric experiment 0.50 0.51 052 could be used to study the kinetics of MOLE FRACTION (XT]0z OR xZr0J 3+ 5+ 3+ 3+ solution of Bi , Nb , La , and Sc ions LEAD ZIRCONATE SINGLE-PHASE REGION-* in PZT and thus the extrinsic non- tLEAD TITANATE SINGLE-PHASE REGION stoichiometry. Results indicated a lead vacancy defect structure created by the Fig. 5. The free energy of mixing PbO and 3+ 5+ 3+ C substitution of Bi , Nb , and La in the either Ti02 or ZrO^ as a function com PZT lattice, and an oxygen vacancy defect position in either the lead titanace or the structure for Sc3*, obeying electro- lead zirconate single-phase region. neutrality rules, and following simple predic CXBL 726-6433) -176- Employment of CAMP PZT crucibles to con TIME. HOURS AT 1150*C trol atmospheric, thermogravimetry may be 10 20 30 40 50 60 70 60 used to study the continuous weight changes CONTINUOUS WEIGHT CHANGE caused by the reaction between any second phase oxide addition with lead zirconate- TIME AT IISO'C 1% LANTHANUM [P6 Ti,ZraO,+ X»3 La»0B] titanate. In this analysis, the numbex of titaniijn and/or zirconiun ions distributed on the B-sito of the perovskite lattice is constant due to the extremely low vapor pressure of these species. However, the lead ion concentration is free to adjust itself by means of equilibration with a surrounding lead oxide atmosphere. Hence, should a cation diffuse into the A-sublattice, some lead ions would be released, keeping the Fig. 6. Continuous weight-change vo. time ratio of A-sites to B-sites at unity. If a (hours) at 1150°C of a pre-treateu cation diffused into the B-sublattice, tte PbTi jZr. 5O3 sample contain ,ig 1 mol i total number of B-sites would be increased, lantnanum oxide (LaOi 5): CO lead oxide and additional lead ions would be obtained atmosphere provided by a stoichiometric from the atmosphere, again holding the PbTi.jZr.5O3 crucible, (b) lead oxiije atmDS- A/B site ratio at unity. This type of phere'pvovided by a low CWP crucible reaction weight-change analysis was first (b-dotted) sample contains ."0 birder. demonstrated by Weston et al.,s who deter (XSL 7010-682DA) mined the substitutional lattice -ite of Fe3+ in lead zirconate-titanate by weighing a sample containing iron oxide before and Preliminary experiments^ conducted by using after a long high temperature firis*. a burial powder of the same composition as the sample confirmed this hypothesis. However, this method may be further refined to reveal the valency of the substituted The present results for 1.0 mol % additions cation and the active charge compensation of MQx to samples of approximately the same mechanism. Also, thermogravimetry has the weight show weight gains which are signifi advantage of continuously recording both cantly larger than reported earlier. This is the weight changes caused by the ionic sub particularly interesting for the case of stitution, including the development of any M32O5 substitution. It is suggested that dominant, though tr~isient, second phases, the refractory niobia particles (initially at at the temperatee of interest. apbCpOO react with the lead oxide obtained from the atmosphere (constant apbo=0.280), Thus, divalent ion substitutions on the forming layers of complex "I'-ad mobates as A-sublat';ice and quatravalent ion substitu equilibrium conditions are approached. No tions OH the B-sublattice of the perovskite sample weight losses were observed until the require only equivalent amounts of PbO lost sample had shown a 1.4 wt %* gain of PbO, At or gained from the original composition. this point the Nb20s-Pb0 phase diagram reveals However, trivalent ion substitutions on the that the composition is S3 mol i PbO, A-sublattice and pentavalent substitutions 17 mol % Nb2&5, which creates about 1.5 vol. % on the B-sublattice require an additional second phase. From the phase diagram it may mechanism to neutralize the charge imbalance be seen that at this composition a liquid created. This neutralization is assumed to phase may have formed. ">e accomplished by the creation of lead vacancies (reduction in the number of lead A similar situation is observed in the ions on lead sites]. bismuth modified system, except that the liquid phase forms from the outset (Bi203 If the CA>-S> crucible is in equilibrium melts at 82S°C). with the PZT sample, all observed weight changes are due to the solution kinetics of Lanthanum and Scandium. This work was the impurity oxides. primarily undertaken to investigate the kinetj-;s of solution of the lanthanum and Bismuth and Niobium. The samples con- scandium oxides. The reaction weight-change taining bismuth and niobium were selected data are presented in Fig. 6(b) and Fig. 7(b) as experimental controls. Crystal chemistry for data representative of 1.0 mol % LaOi.s considerations present a strong case for -aid 1.0 mol % SCOT.,5 respectively. assuming that trivalent bismuth will substi tute for lead and pentavalent niobium will substitute for titanium/zirconium. PbQ t:ain (grams) wt % sample wt (grams)~ -177- 1 I 1 1 ( ( Abstracted from R. L. Holman, Intrinsic and Extrinsic Nonstoichiometry in the lead Zirconate-Tltanate System (Ph.D. thesis), 150 LBL-SSO, July 1972. 3 4 J Sc* /(TiZr) * PREDICTED Present address: Xerox Corporation, Rochester, N.Y. 100 j CONTINUOUS WEIGHT CHANGE 1. B. Jaffe, R. S. Rot,1-, and S. Marzullo, 1 TIME AT 1150 C J. of Research of the National Bureau of bO Standards 55, 239-253 (1955). "j 1% Scandium [PbTi5Zr505 + ,005 ScjO^] 2. R. H. Duhgan, H. M. Bamett, and A. H. Stark, J. Am. Ceram. Soc. 45, 382-388 0 1 1 i 1 1 1 (1962). 0 10 20 30 40 50 60 70 3. K. H. Hardtl and H. Rav, Solid State TIME. HOURS AT M50t Communications 7_, 41-45 (1969). Fig. 7. Continuous .reight-change vs. time 4. D. lee, Sintering Sc and Nb Modified Lead (hours) at 1150°C of a pretreated Zirconate Titanate (M.S. thesis), University PbTi_sSr_503 sample containing 1 raol i of California, Berkeley, UCRL-19188 (1970). scandium'oxide (ScO]^5): (a) lead oxide 5. T. B. Weston, A. H. Webster, and atmosphere provided by a stoichiometric V, M. McNainara, J. Am. Ceram. Soc. 52, 253-257 Pb'fi_5Zr_503 crucible, (b) lead oxide atmos (1969). phere provided by allow CAMP crucible 6. R. B. Atkin, R. L. Holman, and R. M. (b-dotted) sample contains no binder. Fulrath, J. Am. Ceram. Soc. 54, 113-115 (XBL 7010-6824A) (1971). For the lanthanum substitution, no initial weight gains are observed. However, the rate of equilibration is seen to be quite dependent 4. PROCESSING PARAMETERS AND THE PROPERTIES upon the incorporation of a maximum amount OF LITHIUM FERRITE of napthalene to control porosity. Figure 6(b) demonstrates this effect, where Gautam Bandyopadhyay and Richard M. Fulrath (b-dotted) describes the data of a sample which was sintered simultaneously without Lithium ferrite (LiFesOg) possesses some napthalene, and (b-solid) describes the data unique properties, necessary for use in of an identical sample containing approximate computer memory cores. Its processing in ly 7 wt 4 napthalene. In addition, it was volves problems due to the loss of lithium observed that the smaller the sample in and oxygen during the sintering step. The volume, the more rapid the equilibration. purpose of this investigation is to correlate the sintering parameters, stoichiometry, and In either event, the diffusion of La3* the effect of dopants, to the densification, proceeds gradually until an equilibrium is microstructural characteristics, hysteresis reached corresponding to a total La3* loop parameters, and dc resistivity of substitution for Pb2* via a total lead vacancy lithium ferrite. This information will not charge compensation mechanism. only help to control the properties of lithium ferrite, but also would be helpful to under Finally, experiments were conducted to stand the "ferrites" as a whole. characterize the addition of SC2O3. From crystal chemistry considerations it is deduced a. Weight Loss and Chemical Changes that scandium illustrates a trivalent sub stitution on the (Ti,2r)*+ sublattice, but Composition of the starting material is that the substitution occurs with either always an important factor in the processing decreased lead vacancy or increased oxygen of electronic ceramics. The influence of vacancy charge compensation. The data con»)osition on the weight loss of lithium obtained supports this conclusion and indicate ferrite has been determined by thermogravimet- clearly that oxygen vacancies are formed to ric analysis in air. The results tend to neutralize the charge. This is illustrated indicate an increasing oxygen loss with the in Fig. 7. increase of Fe203 content in the composition. This type of material loss is found to be In this case the initial weight gain is enough to cause significant deviation from believed to occur in the manner described the stoichiometry and in some cases led to for niobia. decomposition with the appearance of second phases. All the principle phases that may -178- exist in this ferrite system (e.g., LiFe508, LiFe02, Fe^) have their important X-ray peaks almost at the same angles. But the use of I311/I400 ratio (both planes are for J4.0 the spinel phases) qualitatively indicated the relative amounts of spinel phase and LiFe02 phase present. Figure 1 shows a sharp = 3.0 change in the surface composition of the sintered --rrites on the stoichiometric line. Predominantly spinel phase was observed on the Fe203-rich side of the composition, and 2.0 LiFe02 phase on the lithia-rich side. ! • Bulk resistivity The study of the surface composition was IO also done by using stoichiometric specimens o Surface resistivity and sintering tine, temperature, and packing powder compositions as the variables. The two-step decomposition of lithium ferrite, as originally postulated by Salmon et al.l under somewhat different conditions seems A to explain the results adequately. Further \0 more it seems that Fe203, if present, always tends to go into solid solution with the spinel ferrite as Fe304 and as FeO in the LiFe02 phase. In all the sinteTed ferrites a significant difference has been observed between the bulk and the surface composition. This has been attributed to the different amount of material loss from the two regions. 10' 15 10 5 10 Mole% Li FeO, IO.OF Mole% Fe203 Fig. 2. Dependence of coercive force (He) and dc resistivity on the composition. All the specimens were sintered at 1150°C for 2 hours in 1 atm. O2, using "stoichiometric" packing powder. (XBL 731-2048) b. Processing Parameters and the Properties Most of the sintering runs were done at 1150"C for 2 hours in 1 atm O2 and using stoichiometric packing powder. Some isother mal sintering runs were made for various lengths of time. In some cases packing powder compositions were varied. Sinter densities of Fe203~rich specimens were found to be lower than those of lithia-rich speci mens sintered under identical conditions. This is believed to be due to increased anion vacancy concentration in the lithia-rich compositions. Extensive exaggerated grain growth has Fig. 1. X-ray data for specimens sintered been observed in the lithia-rich specimens at 1150°C for 2 hours in 1 atm 02, using sintered at 1150°C for 2 hours. On the "stoichiometric" packing Dowder. uther hand, very small grains with pores (XBL 731-2050) exclusively on the grain corners have been -179- = 100 sintering have indicated that grain size can also have significant effect on the re sistivity value. In Fig. 3, average size of the largest grains and dc resistivity have been plotted against sintering time. It is clear that discontinuous grain growth is always accompanied hy an ir .rease in resistivity. This is possible if grain boundaries act as the easy conducting path. Oxygen loss is expected to occur preferential ly from or along the grain boundaries. This would result in a higher Fe+2 concentration in the boundaries than that in the bulk and therefore a lower resistivity along the grain boundary. Studies on the effect of packing powder compositions on the sinterability and properties of lithium ferrite are being done. There are indications that significant con trol on the final properties can be achieved by this method. The effect of selective additives like NiO, NiFe204, M11O2, etc. are also being studied with an aim to achieve bet ter control of the magnetic properties. 1. 0. N. Salmon and L. Marcus, J. Am. Ceram. Soc. 43, S49 (I960). 2. ATT.. Stuijts, "Ferrites: Proceedings of the International Conference, July 1970, Japan," p. 108. 10° 10' Time of sintering Fig. 3. Grain size and bulk resistivity 5. DIFFUSION OF GASES IN GLASS data from some isothermal sintering runs. "Stoichiometric" packing powder and 1 atm Joseph Masaryk and Richard M. Fulrath Oo were used in all the sintering runs. (XBL 731-2047] The use of ceramics as either pressure or vacuum envelopes at high temperatures requires knowledge of the ability of gases observed in the Fe203-rich specimens. This to penetrate these materials. From a is in accordance with the results reported practical standpoint fused silica glass is in the literature2 for some other ferrites. of interest because it exhibits the highest overall permeability to gases of any ceramic The variation of coercive force (Hg) and material. Consequently, gaseous diffusion dc resistivity with the composition is shown in fused silica is being studied in an effort in Fig. 2. H^. has been calculated from the to define the precise diffusion mechanisms B-H hysteresis loop. Resistivity was and the effects of thermal history and measured by using a guard ring method which impurities on this process. allowed the measurement of the surface and bulk resistivity separately. Lower He value By far the largest impurity in standard on the lithia-rich side is probably due to commercial silicas is water in the form of larger grains. Variation of resistivity may hydroxy Is (OH), which may be present in the be due to the presence of higher Fe+2 con range from 10 to 1000 ppm, depending on centration in the Fe20j excess compositions method of manufacture. These hydroxyls reduce due to greater oxygen loss. Lower surface the density of the glass at room temperature. resistivity than that of the bulk for an Another factor that affects the dtnsity is individual specimen is expected because of the thermal treatment of the glass which greater oxygen loss from the surface affects the density by increasing it with resulting in higher Fe+2 concentration. increasing stabilization temperature. Experimental runs on the isothermal Very thin (0.07 mm) fused silica diffusion -180- membranes were subjected to a series of treatments designed to independently vary the water content and stabilization temperature. After each treatment the He diffusivity was measured as a function of temperature in the range -78 to 165°C. The activation energies for diffusion of He in fused silica as a nent are: y , cm * aH(diffusion), cal Sample Treatment Density, fj/cc siv i 4905*50.0 lOOO'C, vacuun 2.202 5114*27.5 10QO°C, 1 atm HjO 2.201 I 4967*27.0 1200CC, vacuum 2.204 o _ 5140±50.0 1200 C, 1 atm H20 2.203 I0 These results indicate that water content and stabilization temperature, which affect other properties significantly, do not appreciably affect the diffusion properties of He in fused silica. The small increase in activa Temperature x K>"2 *K tion energy with increasing stabilization Fig. 1. Comparison of experimental diffusi temperature is protably due to the increasing vity data for helium in fused silica with a density-; which decreases the size of the plot of Eq. (1), using the parameters of interstice through which the He atom must d = 2.5 A, v = 6.9x1012 sec-1, move. Even though OH content decreases the v* = 7.8x101-2 sec-l and AH = 6100 cals/mole. density, the activation energy for diffusion (XBL 726-6371A) is increased, because the extra oxygen added to the structure provides a steric hindrance to the movement of He. An interesting observation was noted on the diffusion of He through these thin-walled membranes. Diffusivity data obtained on these membranes over the past two years has con sistently exhibited activation energies about 204 lower than those reported in the literature, which used membranes about 1 mm thick. Consequently, the diffusion of He through a 1 ram membrane was measured in order to distinguish any instrumentation effects. This thicker membrane exhibited activation energies comparable to those reported in the literature. Work is currently under way to determine if surface, compositional, or thermal differences between the membranes Temperature^ I0"2 °K exist. Fig. 2. Variance of the value of the pre- exponential factor, DQ, with temperature of A model describing the diffusivity of the Arrhenius equation for helium diffusion mopAtomic gases in fused silica yields the in fused silica. (XBL 726-6372A) egression •> c> hv . -hv . quency on the atom in the constriction bet kTd' 2 Ce 2KT ' e 2kT J -AHQ/RT ween two sites, v ; and the activation energy "5F (1) for this process, H. Due to the amorphous -h\)* . 2 nature of fused silica it is impossible to ~2kT- accurately calculate the values of these variables. Thus a least squares analysis was performed on the best fit of the model This model, which considers the motion of to diffusivity data of He in SiOo, which the gas atom sis it moves between equilibrium appears in the literature CFig. I). This analysis yields values of 2.5A for d, solubility sites, yields three independent 12 variables: the distance the atom moves 7.80xl0 sec"l forv*, and 6100 cal/mol for between these sites, d; the vibration fre the Mi. These values compare favorably with -181- i . . > ..| i , . i n-T T Use of values of the independent variables determined from the diffusivity and solubility models yields reasonable agreement of the permeability model with data in the literature on He permeation in fused silica (Fig. 3). The permeability model need not be confined strictly to gases in glass, but can be applied - - to any gas which does not dissociate or react and diffuses interstitially in any solid. - f O Svifa, Lea and Frank - ? Modtl 6. RESEARCH PLANS FOR CALENDAR YEAR 1973 Richard M. Fulrath - Processing studies in both lead zirconate .1 titanate and lithium ferrite have established 10 the important process parameters that TertiperoturexlO"' influence the achievement of high density Fig. 3. Comparison of experimental ceramics. During 1973 the major effort will permeability data for helium in fused silica be directed to establishing how intrinsic with a plot of Eq. (2), using the parameters and extrinsic defect structures affect the 2 1 of d = 2.5 A, v •• 7.8xl0l sec" and electronic and magnetic properties of poly- AHfc = 4600. (XBL 726-S375A) crystalline ceramics with controlled density and grain size. those reported in the literature for the The studies on grain boundary structures vibration frequency in a solubility site and in ceramics will concentrate on high density the activation energy calculated using the aluminum oxide. Samples of suitable geometry standard Arrhenius equation. for grain boundary permeation and diffusion studies were fabricated in 1972. The first For the diffusion of non-reactive gases measurements will be to determine the grain in glass, there has been some question as boundary permeation of He in high density to the temperature dependence of the pre- controlled grain size alumina. exponential factor, Do, in the Arrhenius expression. Several authors have argued Sintering studies by high temperature that the expression D = DoTe - AH/RT better scanning electron microscopy will continue describes this process. No theoretical on submicron AljOg. Studies o:i UO2 and justification related to this process has systems containing a liquid phase will be been put forth to justify the insertion of initiated during 1973. a factor of T. Using the model derived in ca n De this work, it can be seen that Do The mechanical properties of glass-ceramics expressed in terms of fundamental parameters will be studied because recent work in this of the material and diffusing species. If group has indicated that the glass matrix one then substitutes the determined values composition can influence the fracture. of d and v* in this expression, Do can be Fracture initiation will be measured to calculated as a function of temperature determine if a microplastic zone is exhibited (Tig. 2]. This yields an essentially by glass. temperature-independent value of Dp from 300°K for He in fused silica. At moderate gas pressures the solution 7. 1972 PUBLICATIONS AND REPORTS of a monatomic gas in glass obeys Henry's law. This yields the relationship that the Permeability=Diffusivity * Solubility. If Richard M. Fulrath and Associate* we combine the diffusivity model of the present work with the solubility model of Journals and books the investigation previous to this one, we obtain the permeability expression 1. J. F. Shackelford, P. L. Studt, and R. M. Fulrath, Solubility of Gases in Glass. 3/2 II. He, Ne, and H2 in Fused Silica. J. Appl. Nr 1 d' h . .-ABl/Kr Phys. 43, 1619-1626 (1972). K = - • ( ') 2 C2) eTkT) 2. R. M. Fulrath, High Temperature Scanning -182- Electron Microscopy, chapter from Electron LBL reports Microscopy and Structure of Materials, edited by U. Thomas, K. M. ttilrath,an d K. M. Fisher 1. J. S. Reed and R. M. Fulrath, Characteri (University of California Press, Berkeley, zation and Sintering Behavior of Barium and California, 1972). Strontium Ferrites, J. Am. Cerain. Soc. (in press) LBL-815, March 1972. 3. R. M. Fulrath, Scanning Electron Micro scopy to 1600°C, chapter in Scanning Electron 2. J. S. Masaryk and R. M. Fulrath, Microscopy/1972 (Part I), Proceedings of the Diffusivity of Gases in Fused Silica--A Fifth Annual Scanning Electron Microscopy Monatomic Model, LBL-847, June 1972. Symposium, IIT Research Institute, Chicago, Illinois (1972). 3. C. B. Shumaker and R. M. Fulrath, Initial Stages of Sintering of Copper and Nickel. 4. R. L. Holman and R. M. Fulrath, Intrinsic Paper for Third International Conference on Nonstoichiometry in Single-Phase Sintering and Related Phenomena at Notre Dams, Indiana, June 5-7, 1972, LBL-848, Pb(Zrn sTin 5)03, J. Am. Ceram. Soc. 55, 192-195 (1972). ~~ June 1972. 4. R. L. Holman, Intrinsic and Extrinsic Nonstoichiometry in the Lead Zirconate Titanate System, (Ph.D. thesis) LBL-880, July 1972. 5. C. B. Shumaker, Initial Stage Sintering of Copper and Nickel, (M.S. thesis) LBL-1110, Aug. 1972. -185- A. THEORETICAL SOLID STATE PHYSICS Marvin L. Cohen, Principal Investigator were to be simulated. These studies yielded information about the origin of the optical Introduction. New insights were gained structure in addition to suggesting schemes into the properties of the amorphous state for systematic studies of the amorphous state. by investigating successively more complex crystalline polytypes of Ge and Si. XPS and UPS have gained renewed popularity in the last few years and it is expected Superconducting transition temperatures that there will be enormous growth in this were related to lattice mode softening, and area in the next few years. Detailed theories limits on the maximum superconducting transi are not available as yet, but it appears tion temperature were explored. fairly certain at this time that the measured spectra reflect the valence band density of states. Our theoretical calculations support this conclusion. In fact, systematic studies 1. SEMICONDUCTORS AND INSULATORS going from covalent to more ionic crystals (Ge, GaAs, ZnSe) show ionic energy gaps which Marvin L. Cohen appear in the experimental spectra and in the theoretical calcuJ ations. in addition, A variety of semiconductors and insulators amorphous films appear to have spectra similar were investigated theoretically,'using the to density of states calculations for the Empirical Pseudopotential Method (EPM). Layer ST-12 systems. It was also possible to show compounds (SnS2 and SnSe2), narrow gap semi some fundamental relations between gaps in conductors (lead salts and CdTe-HgTe alloys], the spectra, energy band splittings, and and complex polytypss of Ce and Si were ionicity parameters. analyzed, in addition to the more popular materials like the III-V and II-VI semi Charge density calculations were done for conductors. The potentials used in these PbTe, HgTe, CdTe, CdxHgi_x Te alloys; cubic calculations were obtained through experi and hexagonal ZnS,'and for model semiconductors mental data input. Optical, X-ray Photo- to study the relation between bonding and electron Spectroscopy (XPS), and Ultraviolet electronic or crystal structure. For the Photoelectron Spectroscopy (UPS) were CdxHgl-x Te case it was possible to explore investigated and used to fit energy levels. the charge distribution of the top valence Calculations of the density of states energy band and bottom conduction band as a function band structures, dielectric functions as a of alloying. This is particularly interesting function of frequency and wave vector, and in this case since the alloy becomes a zero electronic charge density plots were made gap semiconductor near the HgTe end. Model possible by using the EPM scheme. The results pseudopotentials were used in charge density allowed interpretation of the experimental calculations and these showed the role of the data and new insights into the electronic potential in determining the bonding. For the structure of semiconductors and insulators. cubic and hexagonal ZnS case the dependence of bonding on structure was explored and it was For the case of Ge and Si polytypes, shown that for an ideal c/a ratio a dipole investigations were made of the electronic moment exists. It was speculated that changing structure of four modifications of these the c/a ratio might reduce the polarization, crystals: FC-2 (face-centered cubic with two leading to a method for determining c/a atoms per cell), 2H-4 (hexagonal with four ratios. A new scheme was developed for charge atoms per cell), BC-8 (body-centered cubic density calculations, allowing calculations with eight atoms per cell), and ST-12 (simple based only on a few t-points in the Brillouin tetragonal with 12 atoms per cell). It was zone. This scheme considerably reduces the possible to use the same atomic pseudopotential computer time involved in the calculation. for Ce or Si in each of the modifications and to account for the crystal structure For the first time, dielectric functions as change via the structure factor. In this a function of wave vector and frequency were way long-range order was preserved, but short- calculated. The materials studied were Si, range order was continually changed. It was Ge, GaAs and ZnSe. These functions can be found that some of the experimental properties used for electron energy loss studies, bonding of the amorphous state were simulated through studies, and other theoretical investigations the changes in the structure factor. In of semiconductors. particular it was found that the presence of fivefold rings (present only in ST-12) was important if some of the amorphous properties -186- 2. METALS the III-V semiconductors, and the differences between them and the III-V semiconductors Marvin L. Cohen give information about the underlying electron ic structure of these types of materials. They The electronic band structure and charge are also good nonlinear materials. densities, were studied for some metallic sys tems. For the case of aluminum, detailed We also hope to explore the relationship charge density plots showed the electronic between bonding and valence band gaps, do distribution in the unit cell. Fourier trans further studies on Ge and Si polytypes, forms of the charge density were used to photoelectron studies, and further charge calculate the X-ray scattering factors, and density studies. the results were shown to be in good agreement with experiment. Non-local pseudopotentials which were successful in determining the electronic structure of noble metals and 5. 1972 PUBLICATIONS AND REPORTS transition mstals were extended to transition metal compounds. In particular, NbN was Marvin L. Cohen and Associates studied in detail. The band structure and charge density calculation should prove useful Journals and books in studying materials like NbN, which is a * high temperature superconductor. 1. M. L. Cohen and P. K. Anderson, Comments on the Maximum Superconducting Transition Temperature, in Superconductivity in d and f Band Metals, edited by D. H. Douglass 3. SUPERCONDUCTIVITY (Amer. Inst, of Physics, New York, 1972), p.17. Marvin L. Cohen 2. J. Stokes, Y. R. Shen, Y. W. Tsang, Limitations on the maximum superconducting M. L. Cohen, and C. Y. Fong, Wavelength transition temperature were studied. In Modulation Spectra of Single Crystals of particular the dielectric formulation was Silver and Gold, Phys. Letters 38A, 347 (1972). used to show that Umklapp processes were necessary if very high transition temperatures 3. C. Y. Fong and M. L. Cohen, Electronic were to be attained. Energy-Band Structure of SnS2 and SnSe2, Phys. Rev. B 5_, 3095 [1972). The role of phonon softening in super conductivity was also investigated. New sum 4. J. P. Walter and M. L. Cohen, Frequency rules and expressions for the electron-phonon and Wave-Vector-Dependent Dielectric Function interaction parameter were derived. Using for Silicon, Phys. Rev. B £, 3101 C1972). these expressions, the observed transition temperatures are related to phonon softening 5. D. J. Chadi, J. P. Walter, M. L. Cohen, in some materials, and predictions are made Y. Petroff, and M. Balkanski, Reflectivities about materials not yet found to be supercon and Electronic Band Structures of CdTe and ducting. HgTe, Phys. Rev. B5, 3058 (1972). 6. Y. W. Tsang and M. L. Cohen, Electronic Charge Density in PbTe, Solid State Comm. 10, 4. RESEARCH PLANS FOR CALENDAR YEAR 1973 871 (1972). ~ Marvin L. Cohen 7. C. Varea de Alvarez, J. P. Walter, M. L. Cohen, J. Stokes, and Y. R. Shen, We have begun studies of NbC and plan to Wavelength Modulation Spectra and Band use charge density plots to attempt to under Structures of InP and GaP, Phys. Rev. B 6, stand the superconducting properties of this 1412 (1972). crystal and similar crystals. 8. J. D. Joannopoulos and M. L. Cohen, Calculations of NaCl in the zincblende Comparison of the Electronic Structure of and rocksalt structure have been started, Amorphous and Crystalline Polytypes of Ge, and the role of charge distributions in Solid State Comm. 11, 549 (1972). determining crystal structure will be explored. 9. J. D. Joannopoulos and M. L. Cohen, The II-IV-V2 chalcopyrite crystals will be Electronic Structure of Crystalline Polytypes studied with emphasis on ZnGeP2- These and Amorphous Si, Phys. Letters 41A, 71 crystals are interesting since they resemble (1972). -187- * 3. C. Varea de Alvarez. J. P. Walter, 10. M. L. Cohen, Calculations of the R. W. Boyd, and M. L. Cohen, Calculated Band Electronic Structure of Semiconductors, in Structures, Optical Constants and Electronic Proceedings of the 11th International Charge Densities for InAs and InSb, J. Chem. Conference on the Physics of Semiconductors Phys. Solids, (in pri •-) LBL-482, (PWN-Polish Scientific Publishers, Warsaw January 1972. (1972), p. 731. 11. R. A. Pollak, L. Ley, S. Kowalczyk, 4. J. P. Walter, C. Y. Fong, and M. L. Cohen, D. A. Shirley, J. D. Joannopoulos, D. J. Electronic Charge Density of Aluminum, Chadi, and M. L. Cohen, X-Ray Photoemission Solid State Comm., m press (LBL-893), Valence-Band Spectra and Theoretical Valence- June 1972. Band Densities of States for Ge, GaAs, and ZnSe, Phys. Rev. Letters 29, 1103 (1972). 5. D. J. Chadi and M. L. Cohen, Electronic Structure of Hgi-xCdx Te Alloys and Charge 12. C. Y. Fong and M. L. Cohen, Pseudo- Density Using Representative k-Points, Phys. potential Calculations of the Electronic Rev., in press (LBL-1105), July 1972. Structure of a Transition-Metal Conpound- Niobium Nitride, Phys. Rev. B6, 3633 (1972). 6. J. D. Joannopoulos and M. L. Cohen, Electronic Properties of Complex Crystalline 13. S. J. Sramek and M. L. Cohen, Frequency- and Amorphous Phases of Ge and Si: I. and Wave-Vector-Dependent Dielectric Function Density of States and Band Structure, Phys. for Ge, GaAs, and ZnSe, Phvs. Rev. B 6, Rev. in press (LBL-1169), Oct. 1972. 3800 (1972). Other reports * 14. P. B. Allen and M. L. Cohen, Super 1. P. B. Allen and M. L. Cohen, Super conductivity and Phonon Softening, Phys. conductivity and Anomalous Phonon Dispersion Rev. Letters 29_, 1593 (1972). in TaC, in Proceedings of the 15th Interna LBL reports tional Conference on Low Temperature Physics, Boulder, Colorado, 1972 (in press). 1. J. D. Joannopoulos and M. L. Cohen, fe Electronic Charge Densities for ZnS in the 2. W. D. Grobman, D. E. Eastman, and Wurtzite and Zincblende Structures, LBL-1193, M. L. Cohen, A Relationship Between Oct. 1972. Photoemission-Determined Valence Band Gaps in Semiconductors and Insulators and Ionicity 2. S. E. Kohn, P. Y. Yu, Y. Petroff, Parameters. Y. R. Shen, Y. Tsang, and M. L. Cohen, Electronic Band Structure and Optical Proper ties of PbTe, PbSe and PbS, LBL-1132, Not supported by IMRD funds. Aug. 1972. -188- B. FAR INFRARED PROPERTIES OF SOLIDS Paul L, Richavd3, Principal Invegtigator similar behavior to the data shown for the (Co,l*)F2 system, except that cluster defect 1. RESONANCE MODES IN ANTIFERROMAGNETIC modes are not observed. fCo,Mn)F2 AND (Fe,Mn)F2 Simple Ising model calculations of the Paul L. Richards and William E. Tennant mode frequencies agree quite well with the data, as shown in Table I. More detailed The insulating antiferromagnetic iron calculations recently obtained from Oxford group fluorides are perhaps the hest charac are being compared with the results. terized magnetic system. Recent neutron diffraction experiments have attempted to study mode localization in alloys of these materials. Far infrared measurements, which are restricted to wave vector q == 0, are much more sensitive than neutron diffraction and thus provide complementary information. In Fig. 1 we show the frequency and line widths of magnetic resonance modes which ap pear in our far infrared powder spectra of the alloy system (Co,Mn)F2. These modes are identified as antiferromagnetic resonances in pure MnF2 (at 9 cm"1) and in pure C0F2 (at 36 cm"1), and isolated defect modes at 115 cm"1 in nearly pure MnF2 and 27 cm"l in nearly pure C0F2. We also see modes due to clusters of two (32 cm"!) and three 40 1 Mole % MnF2 in Co^ (35.5 cm" ) defects in C0F2 and due to 2 neighbors of isolated defects (67-75 cm"1) Fig. 1. Center frequencies and line widths in C0F2. of resonant modes observed by far infrared Fourier spectroscopy on powder specimens of The alloy system (Fe,Mn)F2 shows very alloys of CoF2 and MnF2. (XBL 7210-7118) Table I. Frequencies of impurity modes. CoF2 FeF Mode Ising Frequency Given By Isine Experiment Is ins Experiment f * 1 1 -1 Hn Defect in Host •^ • H^D- 2SHl'(l) + 8SHl'(2) 28.3 cm W cm" 50 cm 33 cm"1 Heiehbors of Mn Defect 2 1} (2)+S I,(2)f 79 67, 71(2) 79 78-80 "N = "AH" W + ?VH D 72, 75 ** Pairs of Mn Defects Up=H;D-2SHI'(l)+7SHI'(2)+SDID(2) 31.2 32 53 not observed ** Triplets of Mn Dpfpr.rs "E = "in"2s Hr(1) + 6V'V> + 2SDV2) 34.8 35.5 57 not observed Formula and values of 1(1), 1(2), H,, and S taken from R. A, Cowley and W. J. L. Buyers, Rev. Mod. Phys. 44, 406 (1972). Value taken from Cowley and Buyers includes spin-orbit effect of Co neighbors on Co-Mn exchange. Uses experimental value for S..I'(2) computed from defect mode frequency. -189- These experiments answer a long-standing to cool the optics with gas instead of liquid. question: as early as 1962 extra far infrared And finally, a process for coating metal modes were seen in normally pure FeF2 and with polyethylene was developed to reduce the C0F2. We now know them to be due to Mn2+ power emitted by those objects in the beam as a substitutional impurity. that are allowed to remain warm. A first flight is anticipated early in 1973. 1. Samples obtained from Dr. E. Catalano, Lawrence Livermore Laboratory. 2. BALLOON-BORNE FAR INFRARED SPECTROMETER 3. IDEAL PERFORMANCE FROM POINT CONTACT JOSEPHSON JUNCTIONS* John C. Mather, Paul L. Richards and David P. Woody Paul L. Richards and Yuan Taur A far infrared spectrometer has been Point contact superconducting Josephson developed that can be flown with a high tunnel junctions have long been considered altitude balloon for use in astrophysical attractive devices for high frequency mixers studies. The principal features of the and detectors. They are highly nonlinear, system are a liquid-helium cooled polarizing show response beyond ~ 1012 Hz, and have interferometer and a germanium bolometer. low noise. Extensive device calculations The spectrometer is calibrated for absolute based on the resistively shunted junction radiometery with an anticipated accuracy of (RSJ) model have recently shown that better than 0.1°K for a Rayleigh-Jeans Josephson effect microwave mixers can exhibit source. conversion gain. This is in contrast to classical resistive diode mixers in which This spectrometer is being interfaced to the ratio of signal power to intermediate a balloon gondola developed by Professor frequency power [conversion efficiency) must Kinsey Anderson of the Berkeley Space Sciences be less than or equal to unity. Laboratory. The system is designed to make high resolution measurements of the cosmic The RSJ model assumes that the current background radiation in the 6-20 cm"* spectral through an actual Josephson junction is the region. Low resolution measurements in this sum of a Josephson current I = I0 sin(2e/h) x spectral region made from rockets and balloons /V(t)dt and a quasiparticle current I = GV. have yielded contradictory results. In order to test the applicability of the RSJ model to Nb point contact junctions, we Three crucial improvements in the system have measured the dc current-voltage charac were made during the year. First, the teristics of such junctions with and without antenna pattern is now defined by a long applied rf radiation and compared them with polished copper reflecting cone at the calculations2 from the model. Noise in the entrance to the spectrometer. This cone has measuring circuit has a strong influence on no stray reflections as lenses do. Without such curves. When noise is added to the this cone, radiation received from the earth theoretical model, near perfect agreement and the ba" jon would have been stronger than with experiment is obtained, as is shown in the signal of interest. Second, the detector Fig. 1. These results show that calculations has been improved by an order of magnitude of device performance from the RSJ model can by replacing the InSb detector used formerly be expected to be reliable, and that high with a very small Ge bolometer fed by performance mixers should be possible. immersion optics. This was necessary because of the very small signal entering the spectrometer (only 2.5 x in-11 w over the entire bandwidth of the 2.7° background radiation). Third, the cryogenic efficiency 1. A portion of this research was supported has been improved in several ways, allowing by the U.S. Office of Naval Research and the sufficient time to take data in flight before U.S. Army Research Office, Durham. the liquid helium boils away. A superfluid 2. Computer calculations by F. Auracher, pump was developed to circulate the liquid Department of Electrical Engineering, U.C., on comuand. Gas flow systems were devised Berkeley. -190- polarizing interferometer. This instrument takes advantage of the the superior properties of the free-standing wire grid polarizer as a beam-splitting element as compared to the thin Mylar dielectric beam solitters normally used in Michelson interferometers in the far infrared. The efficiency of dielectric beam splitters is strongly frequency-dependent due to interference effects, and has a peak value of only 601. The free standing wire grid polarizing beam splitters, on the other hand, have a constant efficiency of nearly 1004 up to a cutoff frequency which is determined by the wire spacing. Construction of the polarizing interferometer lias now been completed and it is in general labora tory use over the frequency range from 1 to SO cm"1. The output of a Fourier transform spectrom eter is an interference pattern (or interferogram) which is the Fourier transform of the desired spectrum. Hence, the basic tool of data analysis is the Fourier trans form. In order to provide a quasi-real-time data analysis capability, our data collection system is interfaced with a PDP 11/20 minicomputer. Data acquisition is computer controlled. The data can be stored in active memory and/or Fourier transformed. Using a fast Fourier transform algorithm R F current Irf /Ic developed at the Argoime National Laboratory, 1000 data points can be transformed in about Fig. 1(a) Experimental and theoretical I-V 3 seconds. curves for a Nb point contact with T = 4.2°K and normalized frequency fl = 0.16 at 36 GHz. Other mathematical operations include The data are shown as a heavy solid line. apodization (correction for the finite length The thin solid line is the prediction of the of the dan), coding of repeated data runs RSJ model without noise and the thin dashed (for improvement of the signal-to-noise line is the prediction of the RSJ model with ratio), convolutional phase correction 100K thermal noise. (correction for sampling errors), division (b) Experimental and theoretical dependence of one spectrum by another (to remove the of the zeroth and first step heights on rf spectral profile of the spectrometer from current. The theory is fitted to the data the spectrum), and calculation of the at the two marked calibration points. This logarithm of the spectrum (for computation two-point calibration is used to fit all of the absorption coefficient). Either the curves in Fig. 2(a)-(b). The resistance value data or the transformed spectrum can be stored deduced is the same «s that required to fit in active memory, displayed on an oscilloscope, the static I-V curve in the low voltage region. plotted on an x-y recorder, punched on paper (XBL 728-3744) tape for permanent storage, listed on the teletype, or punched on cards for analysis on a general purpose computer. 4. INSTRUMENTATION FOR FOURIER TRANSFORM SPECTROSCOPY Richard L. rtjrrbachan d Paul L. Richards 5. FAR INFRARED STUDIES OF y-IRRADIATED RUBY Development of instrumentation for Fourier Richard L. Aurbach and Paul L. Richards transform spectroscopy proceeded along two lines: construction and testing of a new The formation of color centers in irradi Fourier transform spectrometer and computer ated ruby has been studied by numerous programming for real-time data analysis. researchers for over ten years. This work has been stimulated by the importance of Last year construction was begun on a ruby as a laser material, particularly since -191- ultraviolet radiation from laser flash lamps 6. GENERATION OF TUNABLE FAR INFRARED has sufficient eneTgy to form these centers. RADIATION BY MIXING TOO DYE LASERS One model which has been proposed to Kei-hsiung Yang, James Morris, Paul L. explain these color center effects involves Richards, and Y. Ron Shen change transfer reactions which lead to the formation of Cr2+ and Cr^+ ions in the crystal. In order to make far infrared difference EPR experiments predict that the Cr*+ ion frequency generation into a practical in ruby should absorb radiation in the 7 cm"1 spectroscopic tool, we have developed the region. Thermal conductivity measurements dual-beam one-cavity dye laser source, suggest that Cr2+ may absorb radiation near shown in Fig. 1. Wher. a Glan-Thomson prism 6 and 65 cm"1. is inserted into the cavity of a dye laser, the horizontal and vertical polarizations We have measured the far infrared trans are sent in separate directions. Diffraction mission of an 18-cm iong ruby rod* over the gratings can then be used as frequency selec frequency range from 3 to 400 cm"1. The rod tive Teflectors to return the beams back has been measured after y-irradiation from to the dye cell. By turning the gratings a °nCo source, after bleaching with white independently, the laser can oscillate light, and after annealing at high tempera simultaneously with orthogonal polarizations ture. The Y"irradiated sample shows broad at two different frequencies, ""or a ruhy background absorption due to damage-induced pumped dye laser operating on L.rC iodide phonon transitions and also weak absorption in dimethyl sulfoxide, the two lasTS can be lines at 6.4, 7.5, and 59 cm"1, Work is in varied over the bandwidth from 810i to 8400A. progress to positively identify these lines. The laser output can be mixed in noi-linear crystals to generate tunable far infrared ra A second sample1 which contains 100 ppm Ba diation over the frequency range from 5 to ] s also been studied. It is too short beyond 190 cm"1. Many nonlinear crystals such (1.9 cm) to show the weak lines found in the first sample, but does show the broad back ground effects as well as a strong radiation -1 induced impurity mode at 150 cm . Fig. 1. Schematic diagram of the dual-beam one-cavity dye laser used for generation of the tunable far infrared radiation. The 1. Obtained from R. L. Forman, NBS, symbol © indicates that the optical polariza Washington, D.C. tion is perpendicular to the plane of the diagram, t indicates that the polarization is parallel to the plane of the diagram, The nonlinear mixing crystal is mounted on a rotaiable disk. The difference frequency signal at (U1-W2 is normalized by the sum frequency signal at 0)I+ID2. fXBi- 7^1-5600) ^r^ GRATING I GRATING I jf>Z PHOroCCNDUCTIVE DYE CELL CRYSTALS DETECTOR RUBY Hi-ofe. Ge: Go r-*-*SH ot 4.2"K GLAN THOMSON POLARIZER W LASER CAVITY G0A,V- ^j- PHOTO MULTIPLIER -192- as LiNbC^, ZnO, CdS, CdSe, and ZnS have been be measured. A search will be made for investigated for this purpose. For genera effects related to localization of the modes ting far infrared radiation at frequencies observed. It is anticipated that construction below 511 cm"*, LiNbOj is the best material will be completed in early 1973 on the tried. For frequencies higher than 50 cm'l, balloon-borne radiometer for cosmic background ZnO (which can be collinearly phase-matched measurements. At least one balloon flight as high as 190 cm"1!) is more suitable. This should be carried out during FY 1974. Studies dual-beam one-cavity dye laser scheme has of the far infrared electron spin resonance also been tested successfully with rhodiamine of samples of AI2O3 containing Cr2+ and Cr4+ 6G dye and a flash lamp pump. The flash- will be carried out in order to interpret lamp-pumped laser can be pulsed repetitively, data on radiation damaged ruby. Development thus greatly increasing the usefulness of of the dye laser difference frequency source the method. will be continued and further studies of the nonlinear coefficients of optical materials in the phonon frequency range will be carried out. It is anticipated that a program to 7. NONLINEAR OPTICAL SPECTROSCOPY OF x^ explore the far infrared properties of liquid IN LiNb03 crystals will be initiated. James Morris, Kei-hsiung Yang, Paul L. Richards, and Y. Ron Shen 9. 1972 PUBLICATIONS AND REPORTS The dual-beam one-cavity dye laser scheme described above has been used to investigate Paul L. Richards and Associates the dispersion of the nonlinear susceptibility a n x(2)(- U1-U2), (u>l, ~ 2) i LiNb03. The far Journals and book infrared difference frequency power at OJI-O)2 is measured as a function of the dye laser 1. J. C. Mather, M. W. Werner, and frequency u>2, "i*n the laser frequency u>i kept constant. The absorption coefficient P. L. Richards, A Search for Spectral Features at 0)1-0)2 must be known in order to relate in the Sub-Millimeter Background Radiation, the measured power to x^- Absorption Bull. Am. Phys. Soc. (Ser. II) 17, 123 (1972). coefficients as large as 1500 curl (which 2. Y. Taur and P. L. Richards, The Josephson can not be measured by linear spectroscopy] Junction as a Phase Locked Heterodyne are obtained by varying the crystal Detector, Bull. Am. Phys. Soc. (Ser. II) 17, orientations and measuring the width of the phase matching curve. In order to normalize 45 (1972). laser beam fluctuations, the signal at the sum frequency, u>i+a)2, was generated by the 3. K. H. Yang, P. I.. Richards, and reflection method from the surface of GaAs. Y. R. Shen, Theoretical Study of Coherent The expected enhancement of x^) due to Phonon Generation by Optical Mixing in a polariton effects has been observed when the One-Dimensional Superlattice, Bull. Am. Phys. difference frequency u>i-ca2 is close to a Soc. (Ser. II) 17, 47 (1972). transverse optical phonon mode in LiNb03< We also olan to measure the dispersion of 4. P. L. Richards, Magnetic Suspension and (2) in CdS. Propulsion Systems for High-Speed Transpor x tation, AIP Conference Proceedings: Magnetism and Magnetic Materials 1971, edited by C. D. Graham, Jr., and J. J. Rhyne (American Institute of Physics, New York, 1972), p. 935.+ 8. RESEARCH PLANS i !R CALENDAR YEAR 1973 5. P. L. Richards and M. Tinkham, Magnetic Paul L. 'Richards Suspension and Propulsion Systems for High- Speed Transportation, J. Appl. Phys. 4J5, With the exception of the Josephson 2680 (1972).+ junction work which is now being supported by the Office of Naval Research and the LBL and other reports interferometer construction, which is now completed, all of the projects discussed 1. K. H. Yang, P. L. Richards, and will be continusd. Exploration of the zero Y. R. Shen, Coherent Phonon Generation by field far infrared resonance modes in the Optical Mixing in a One-Dimensional Super- iron group fluorides will continue on the lattice, LBL-889, June 1972. (Fe,Co)F2 and (Zn,Mn)F2 systems, and Zeeman splittings of the modes- already discovered 2. Y. Taur, P. L. Richards, and F. Auracher, in the (Fe,Mn)F2 and (Co,Mn)F2 systems will The Application of the Shunted Junction Model -193- to Point Contact Josephson Junctions, in 6. P. L. Richards, Shielding of the Magnetic Proceedings of the Thirteenth International Field Onboard Superconducting Trains (1972 Conference on Low TomperatUTe Physics (LT-DV), Materials Research Council Report).+ Boulder, 1972 [to bo published); LBL-1136, Aug. 1972. 7. P. L. Richards and M. Tinkham, Analysis of Ride Quality in High Speed Trains, (1972 i. P. L. Richards, F. Auracncr. and T. Van Materials Research Council Report).''' Duzer, Millimeter and Submillimeter V.'ay» Detection and Mixing with Superconducting Weak Links, (to be published). 4. F. Auracher, P. L. Richards, and Sponsored by ARPA (Advanced Research Project G. I. Rochlin, Observability of Quasiparticie- Association). Pair Interference Current in Superconducting ft Weak Links, (to be published); LBL-1197,* Sponsored by the ONR (Office of Naval November 1972. Research). Partially sponsored by the ONR (Office of 5. B. Enders, P. L. Richards, and W. E. Naval Research). Tennant, Antiferromagnetic Resonance Modes in (Co,Mn)F2 and (Fe,Mn)F2, Proceedings of the 18th Annual Conference on Magnetism a^d Magnetic Materials, Uenver, ColoV, November 2iR-'ecember l, ia/2 (to be published); LBL-1391, Oct. 1972. -194- SUPERCONDUCTIVITY Gene I. Boahlin, Principal Investigator 2. BEHAVIOR OF WEAK-LINK DEVICES NEAR THE JUNCTION-LIKE TO BULK-LIKE TRANSITION 1. SHUNTED JOSEPHSON JUNCTIONS: EXPERIMENTS AND MODEL Yeong-du Song and Gene I. Rochlin Paul K. Hansma and Gene I. Rochlin The dc characteristics of small thin-film tin microbridges both with and without applied Our previous work on externally shunted rf field has been studied at temperatures Josephson junctions has been extended to below the superconducting transition tempera junctions constructed in a non-inductive ture of tin. When the temperature and configuration as shown in Fig. 1. By impurity-dependent coherence length, £ is simplifying the sample geometry in this way greater than the bridge length, L, the tem possible complications due to the inductance perature dependence of the dc supercurrent of the open loop shunts previously used have agrees with predicted behavior of ideal been eliminated, and a very precise fit to Josephson tunnel junctions. In this regime, the McCumber shunted junction m>del can be the rf-induced ac steps on the junction made. In order to further explore the nature characteristics occur only at harmonics of the of the model an eyact mechanical analog, a rf frequency, e.g., at voltages V = nhu/ze, driven, damped pendulum, was constructed and and the power dependence of the steps is in its characteristics examined. By plotting good agreement with Josephson ac effect theory. the mechanical analogs of the junction I-V When the temperature is decreased or the characteristics, the time-averaged rotation coherence length otherwise shortened so that rate vs. applied torque for the pendulum £ < L, the rf sensitivity is markedly system, the McCumber solutions can be examined decreased ar.J vast numbers of small steps in a detailed way to gain physical insight appear at subuarmonic and fractional harmonic into the behavior of weak-link devices. voltages mV = nftw/ze. This clearly shows that an ac Josephson effect is observable only for Josephson analog systems when the coherence length is less than or comparable to the weak link of the system, and supports recent arguments that the ac Josephson effect in liquid helium has, in fact, never been observed, as L » £ for the geometries used in these experiments. The frequency and temperature dependence of the Dayem effect [enhancement of the zero- voltage step with applied microwave power) has led us to a new theory of this effect based on increased quasiparticle lifetimes very near the transition. Our data very clearly show that the usual explanation of phase-fluctuation stabilization by the rf field is untenable for our samples. 3. INDIUM-BARRIER-INDIUM TUNNELING JUNCTIONS Michael D. Jack and Gene I. Rochlin cm- h We have successfully produced high-quality In-In and In-Pb superconducting tunnel 0Tin junctions by a completely new junction Silver fabrication technique--glow discharge Fig. 1. A non-inductive externally shunted dissociation of a liquid fluorocarbon. By Josephson junction. The outer two junctions this method we have been able to controllably are shunted by Sn-Ag tunnel junctions, since vary the junction resistance over 12 orders the Ag is deposited first. The center junc- of magnitude, and to produce the first In-In ion is left unshunted to monitor junction tunnel junctions ever reported. The second quality. ptBL 721-2208) derivative of the In-In characteristic, as -195- 5. OBSERVABILITY OF QUASIPARTICLE-PAIR INTERFERENCE CURRENT IN SUPERCONDUCTING WEAK LINKS Franz Auracher, Paul L. Richards, and Gene I. Rochlin The effect of the phase-dependent pair- quasiparticle interference conductance, Gl cos 4. CHARACTERIZATION OF TUNNELING BARRIERS Duncan E. McBride and Gene I. Rochlin We have obtained excellent agreement be tween the experimentally measured current- voltage characteristics of metal-insulator- metal tunnel junctions and a model calculation based on a trapezoidal barrier. By comparing the junction characteristic, over many orders of magnitude of current, with the results of the calculation, we are able to determine not only s y = 3i/r,0. (XBL 7211-7182) -196- for the effect, and the negative evidence is 2. Y. Song, and G. I. Rochlin, AC Josephson thus inconclusive, since the ratio of the Effect and Temperature Dependence of the interference to the normal conductance may Critical Current in Superconducting depend on the magnitude of the Josephson Microbridges, APS Meeting, San Francisco, dc current. Calif., January 1972. 3. P. K. Hansma and G. I. Rochlin, Observa tion of Phonon Structure in the Tunneling 6. RESEARCH PLANS FOR CALENDAR YEAR 1973 Density of States of Superconducting Vanadium, APS Meeting, San Francisco, Calif., Gene I. Rochlin January 1972. The projects listed above will be completed, LBL reports and the search for the pair-quasiparticle interference term will be continued. 1. G. I. Rochlin and P. K. Hansma, Inexpensive Model of a Josephson Weak Link, LBL-1185, October 1972; submitted to Am. J. Phys. 7. 1972 PUBLICATIONS AND REPORTS 2. Paul Kenneth Hansma, Externally Shunted Gene I. Rochlin and Associates Josephson Junctions (Ph.D. thesis), LBL-498, March 1972. Journals and books 3. P. K. Hansma and G. I. Rochlin, Josephson 1. Yeong-du Song and G. I. Rochlin, Transi Weak Links: Two Models, to be published in: tion from Bulk Behavior to Josephson-Like Proceedings of the 13th International Behavior in Superconducting Microbridges, Conference on Low Temperature Physics, Phys. Rev. Letters 29, 416 (1972)(LBL-422). Boulder. Colorado, Aug. 21-25, 1972; LBL-851, August 1972. 2. P. K. Hansma and G. I, Rochlin, Josephson Weak Links: Shunted Junction and Mechanical 4. M. P. Jack and G. I. Rochlin, Indium- Model Results, J. Appl. Phys. 43, [11], 4721 Barrier -Indium Tunneling Junctions, ibid.; (1972)(LBL-829). LBL-1150, August 1972. Papers presented 1. G. I. Rochlin, Externally Shunted Josephson Junction, invited paper, APS Meeting, San Francisco, Calif., January 1972. -197- John Clarke, Principal Inoeatigatov This result is in good agreement with the theoretical estimate. i. NON-EQUILIBRIUM EFFECTS IN SUPERCONDUCTORS This experiment is the first to demonstrate John Clarke the existence of such non-equilibrium potential differences in superconductors. We have observed a new effect in super There are several physical situations in conductors which are in non-equilibrium. which these effects are important: an example The experimental configuration is as follows. is given in section 5. A strip of aluminum is evaporated onto a glass microscope slide and oxidized, and a strip of tin evaporated across the aluminum 10 1 1 i to form a tunnel junction. The junction is i then masked off with varnish, except for a • Experiment small window in the middle of the junction. 1 A strip of copper is evaporated diagonally — A" across the window, to form a probe. Any x A"8 voltage generated across the probe and the tin strip can be measured with a super conducting voltmeter (SLUG). X With the tin in the supercond'icting state, a current I is passed between the aluminum and tin strips, injecting quasiparticles into X P - the tin side of the junction and removing > pairs from it. The tin is therefore in a non-equilibrium situation, with an excess population of quasiparticles and a depleted population of pairs. Furthermore, there is 1 an imbalance in the quasiparticle populations above and below the Fermi surface of the tin. en 4 Under these circumstances, there is a dif C>? ference between the chemical potential of X I the pairs, measured by the tin strip, and the quasiparticle voltage, measured by the copper strip. It can be shown theoretically x 4 that for higher injection voltages (well above the gap A of the tin) the voltage dif x -/ ference is • M y )L^-^» » IT„ V = 2e fiNW*N S i i , 12 3 4 where TQ is the relaxation time for the quasi particle imbalance, a the volume of the tin Temperature (•«) overlaying the aluminum, N(0) the density of 1 1 states in the tin, and gNS the low voltage Fig. 1. Plot of £ = VGgNs " against tem tunneling conductance between the tin and per?, ture, averaged over all samples. the copper. Figure 1 shows the quantity (XBL 731-2042) ? = vs^r1 plotted against temperature. The experimental 2. ELECTRON FOCUSING IN BISMUTH: A NEW points are a good fit to a curve proportional METHOD OF MEASURING THE CALIPER OF A FERMI to A"I, from which we deduce SURFACE 4 x 10"10 A(0)/A(T) John Clarke and Louis Schwartzkopf We have observed electron focusing in bismuth in the presence of magnetic fields of -198- -50fl of the two different combs, and a maximum _ ' i when they are focused between teeth of the — same comb. From the spacing of the teeth and the values of H at which the extrema — Normal occur, the caliper of the Fermi surface in metal a plane perpendicular to the sample surface can be determined. ' Superconducting film -Insulotor Single-crystal samples were grown from ~<&z$z%zzz!5zz2za&- 99.999 + % bismuth, and spark-cut to the r-t proper orientation. The spark-cut surface J--I00/A was chemically polished, and cleaned in a glow discharge. The tin combs were then .Teeth of comb evaporated onto the cleaned surface without exposing the sample to the atmosphere. The T*™ resistance between the combs was measured by a Josephson junction device (SLUG) in an ® H Normal automatic self-balancing circuit. "metal Electron orbits Figure IB shows our results from a sample in which the plane of the combs is perpendicular to the trigonal axis. The observed positions of the major maximum and minimum in R(H) agrees very well with the Fig. 1. Comb structure: A--plan view, value HQ calculated from the known Fermi showing interlocking combs; B--side view, surface parameters (see Fig. 2). In fact, showing electron orbits. (XBL 7112-4954) we believe that our method may provide the most accurate way of directly measuring the caliper of a Fermi surface. The extrema in R(H) at Ho/2, Ho/3, etc., correspond to electron focusing between more distant teeth of the combs. We plan to complete the work on bismuth by obtaining data in the other two crystal- lographic orientations. We expect to see the hole orbits in these cases. We hope to be able to see similar focusing effects be tween just two superconducting strips. In addition, we plan to study other materials, such as copper, with this technique. 3. AN IMPROVED DC SQUID John Clarke and James B. Kruger For a number of years we have used double Josephson junction devices (dc SQUIDS) to Fig. 2. Resistance between combs evaporated measure tiny magnetic fields and voltages. on the trigonal face of a bismuth crystal as During the past year we have made two improve a function of magnetic field. Maxima are ments which have greatly enhanced both thev visible at Hn/2 and HQ/4 , and minima at Hp, sensitivity and the ease of operation of our and 1^/3. (XBL 731-5625) devices. The first improvement is to couple the a few gauss. Two interlocking superconducting ac SQUID into a resonant circuit, as indicated combs are evaporated onto the surface of the in Fig. 1. The SQUID is biased at a non-zero crystal, (Fig. 1A) and the resistance between voltage by means of a dc current IQ, and a the two combs R(H) measured at liquid helium small dc magnetic field is applied at the temperature as a function of a magnetic ^ield resonant frequency of the tank circuit, v. parallel to the teeth. R(H) has a minimum If the dc magnetic flux is ncJg/2, where $0 when the electrons are focused between teeth is the flux quantum, the signal developed -199- use, and we feel that this configuration will prove to be the most sensitive of the SQUIDS. 4. A JOSEPHSON JUNCTION BOLOMETER John Clarke and Gary I. Hoffer Last year we reported on the use of the superconductor-normal metal-superconductor Josephson junction as a highly sensitive infrared bolometer element. At that time we had demonstrated the feasibility of such a device. In the past year we have con centrated on improving the reliability and optimizing the sensitivity of the bolometer system. We can now make junctions reproduci- bly on (low heat capacity) sapphire sub Fig. 1. DC SQUID connected in resonant strates, and junctions stored and recycled circuit. (XBL 731-5620) over periods of up to eight months have shown no change in electric?1 properties. The chief limitation on the sensitivity of the across the tank circuit is at a frequency bolometer has been noise in the SLUG that we 2v, and is alncst entirely rejected by the have used to measure the voltage across the tank circuit. If the dc flux is (n + 1/2) junction. We have recently acquired an i|>o/2, the signal is at a frequency v, and rf SQUID which, when used in our bolometer is amplified by the Q of the tank circuit, system, provides a voltage sensitivity 100 typically 100. This signal is detected with times better than that of the SLUG. a lock-in detector, and the output fed back into a coil coupled to the SQUID to provide We have investigated the use of thin negative feedback. The device is therefore bismuth films as an infrared absorbing the null sensor in the feedback system: this coating for our bolometers. The infrared procedure overcomes almost all sources of transmission in the frequency range 0 to 50 1/f noise. cm-1 was measured, and it appears that a relative absorption close to the theoretical The voltage developed across the tank maximum of 3/4 can be obtained with ~ 300 A circuit when the dc flux is changed from bismuth films on sapphire. We plan to test n 1 1 1 • • - 1 1 I.OS Sn-Ir -Sn /. SPEC. 9E 1.07 - THICKNESS = 51.t > |- 106 - O JATA THEORY & ' 10 105 T0 " l-l'IO" SEC. K, »/R, r0 • i.3>Kr SEc 0 10 . . I.5.I0" SEC. - To 1.04 A i - 1.03 r 102 - *S - ^ 1.01 -s-2-Sr 100 - 1 1 • 0.5 0.6 0.7 0.8 0.9 1.0 VTC Fig. 1. Resistance normalized to bulk value versus reduced temperature T/Tc, where Tc is the critical temperature for tin. QCBL 731-5622) of S. This has indeed been shown to be the case by making sandwiches which are identical to [A(T)]-l/2, where A(T) is the energy gap in all respects except for the S. In parti in S at the temperature T. A computer fit cular, using a known superconductor (Ir) as of the data to the theory (see Fig. 1) then N, both Pb and Sn (s) gave the accepted value determined the branch-crossing relaxation for Ir of N(0)V = 0.11. tins TQ = 1.3 x 10-10 sec, in excellent agreement with the theory. Specimens were then made of Pb-Rh-Pb, Pb-Ag-Pb, Pb-Cu-Pb, and Pb-Mg-Pb. The first three types were run in our dilution refrigerator from 18 mK to 7K. A careful 6. A SEARCH FOR SUPERCONDUCTIVITY IN analysis of the data indicated that at the DEGENERATE SEMICONDUCTORS lowest temperatures it is necessary to include the effect of the self-field of the John Clarke, Richard Dalven, Stuart M. Freake, measuring current on the resistance. Efforts James B. Kruger, and Michael L. Rappaport at calculating this effect have provided improved fits to the data and have suggested The change of susceptibility of a semi alternative measuring schemes. conductor was used to detect the super conducting transition down to 18 mK. Since The Pb-Mg-Pb specimens indicated a dirty only three semiconducting superconductors SN interface due to the reactivity of the Mg. are known (GeTe, SnTe, SrTi03), the discovery To eliminate this problem, a new sputter-etch of additional ones should provide new insight system was designed and built and will soon into the relationship between the normal be in operation. state properties (band structure, carrier concentration) and the mechanisms (electron- In addition, the resistance of the Sn-Ir-Sn phonon, electron-exciton) for superconducti specimens near the Tc of Sn was analyzed as a vity. non-equilibrium problem of quasiparticle injection into a superconductor (see The samples tested were AuTe2, GaP, SnSe, section 1). According to the Clarke-Tinkham GaSe, o-Sn, HgSe, PbSe, and Ge. From the theory, the quasiparticles should travel magnetization curves of ct-Sn and HgSe, it dissipatively into S a distance proportional appeared that part of these specimens went -201- superconducting. However, since both S-Sn films where the lower carrier concentration and Hg are by themselves superconductors, (ID17/ -m3 as opposed to 1022/cm3) would inhomogeneous samples could account for the allow larger dimensions and greater noise. results. With bismuth the 1/f noise was often easily observable but usually only when the samples The single crystal of high purity natural carried currents close to their burnout (isotope concentration) Cd included in each value. Samples exhibiting noise showed run to be sure the apparatus was working had definite visible deterioration after an hour ± Tc = 0.S20 0.001 K, in agreement with or so of use. Recent experiments with an recently published values, but differing improved configuration to eliminate noise considerably from Roberts'* values due to pickup and the current supply showed (0.54-0.56 K). that the 1/f noise power in bismuth was at least three orders of magnitude below that predicted by the formula above, provided the sample was operated at currents well below the * burnout level. Also, with the improved Handbook of Chemistry and Physics, 48th configuration, 1/f noise of the predicted edition, edited by R. C. Keast, 1967. magnitude was not observed in gold films. In conjunction with the experiments, a 7. 1/f NOISE IN THIN FIIWS theoretical study of possible 1/f noise theories was begun. All present theories are John Clarke and Richard F. Voss inadequate. Non-equilibrium statistical mechanics and nonlinear theories seem to be A study was undertaken to determine the the most promising. A computer simulation physical mechanism of 1/f noise through an of Brownian motion with a cubic restoring investigation of the simplest and best under force gives c. 1/f spectrum over at least stood system purported to exhibit 1/f noise, several decades of frequency. However, that of thin metal films. Although most experimentally there seems to be little investigators have observed 1/f noise in metal correlation between nonlinearity in the films only when the metal was non-uniformly current-voltage characteristic and the meas deposited in the form of small "islands," ured 1/f noise. one group has reported the noise in continuous gold forms and suggests that metal films and The theoretical work will be continued. other systems produce 1/f noise according An attempt will be made to set a more quanti to: tative upper limit to the 1/f noise in metal samples, and to correlate the onset of /AV\2 _ 2 x 10"3 if observable noise with irreversible changes in the system. where N » total number of charge carriers in the sample. For pure metals one then 8. QUANTUM TRANSITIONS IN A DC SQUID requires cross-sectional areas of order 10"8 cm2 or smaller to observe the 1/f noise John Clarke and James L. Paterson above amplifier background noise. We have observed quantum transitions in A technique was developed for cutting the dc SQUIDS at currents below the critical samples from thin films evaporated onto current Ic for the case of an asymmetric glass slides, using a diamond knife and a current feed in zero applied field (Fig- 1A). two-dimensional micromanipulator. The The Sn-Sn oxide-Sn junctions are shunted by technique proved extremely versatile, allowing silver disks. A tin ground plane (not the construction of samples in various shown), insulated from the SQUID by 500 A of geometries with multiple lead connections' and SiO, was used to decrease the loop inductance. dimensions ranging from a few microns to several centimeters. Consider the case in which the junctions are identical with currents of mfo/L each, The initial measurements on gold films where L is the loop inductance, ife iissume were limited by electrical pickup from the that the SQUID quantizes flux exactly in the current source and only those strips with limit m » 1. The current I divides non-uniform cuts containing "necked down" inversely as the inductances of the current regions produced observable noise. Since paths (Fig. IB). For I » Lj, the junction this was not the simple reproducible system inductance, essentially all cs: I flows desired, an attempt was made to use bismuth through junction A. When T^ reaches IE$O/L, -202- Fig. 2. Current (vertical, 500u A/cm) versus voltage (horizontal, SOu V/cm) for asymmetric SQUID. T = 1.3 K. (XBB 7212-6310) '~nd-\. 1 i-i For a SQUID with hysteresis in the I-V characteristic there is the possibility that these voltage pulses will trigger a transition from the superconducting to the finite voltage state. Figure 2 shows the I-V trace swept at ~ 100 Hz for a SQUID at 1.3 K with I "» 1.3 mA. For I > 0.7 mA there exists (b) c a stable state at finite voltage. Several transitions are seen, separated in current Fig. 1. (a) SQUID consisting of two Sn-SnO*- by 101 ± 10 uA. The transitions for positive Sn tunnel junctions with Ag shunts. A 500 A I appear to carry approximately equal weight layer of SiO and a Sn ground plane are (there can be only one transition per current evaporated over the structure. sweep). The measured value of V(t)dt = n* f 0 9. RESEARCH PLANS FOR CALENDAR YEAR 1973 Consider the case wheTe n is always unity. The flux quantum in the loop is produced by John Clarke a circulating current These voltage pulses for I < Ic have not obtain the variation of the relaxation time, been reported previously. TQ, as a function of energy. These results -203- can then be applied to related effects in the discrete energy levels of quasiparticles non-equilibrium superconductors. "trapped" in the normal layer. The new technique for determining the caliper of Fermi surfaces will be applied to bismuth in different orientations, and 10. 1972 PUBLICATIONS AND REPORTS perhaps also to copper. From the results we expect to be able to make accurate John Clarke and Associates estimates of the various calipers. We hope to be able to calculate the amplitude of the Journals and books , oscillations theoretically. 1. J. Clarke, Experimental Observation of The toroidal SQUID and the resonant Pair-Quasiparticle Potential Difference in detection technique will be further developed, Non-Equilibrium Superconductors, Phys. Rev. and incorporated in a feedback system. The Letters 28, 1363 (1972) (L2L-8i7). ultimate noise limitations of this system will be determined. We hope to make this 2. M. Tinkham and J. Clarke, Theory of Pair- system our basic laboratory measuring Quasiparticle Potential Difference in Nor- instrument, as a more sensitive replacement Equilibrium Superconductors, Phyi. Rev. for our present devices. Letters 28_, 1366 (1972) (LBL-818). The resolution of the Josephson junction 3. J. Clarke and S. M. Freake, Super far infrared bolometer will be measured. If conducting Fuuntain Effect, Phys. Rev. Letters the sensitivity meets our expectations, the 29, 588 (1972) (LBL-897). bolometer will be developed into a hardened device for a number of applications. We 4 J. Clarke, S. M. Freake, M. L. Rappaport, expect to collaborate with Professor Richards and T. L. Thorp, A New Exp. rimental Technique in these applications. for Determining the BCS Interaction Parameter in Normal Metals, Solid State Communications tore data will be obtained on the resist 11, 689 (1972) (LBL-457). ance of SNS junctions at low temperatures. From these data we hope to be able to make 5. J. Clarke, Low Frequency Applications of quantitative estimates of the transition Superconducting Quantum Interference Devices temperatures of a number of materials, for (invited review article for Proc. IEEE) example, Cu, Ag, Mg, and Rh. (LBL-1112). More degenerate semiconductors will be 6. J. Clarke, Low Irequency Measurements with tested to see if they become superconducting Josephson Devices, invited paper at the at low temperatures. The transition tempera Annapolis Applied Superconductivity Conference, tures of any materials which are super May 1-3, 1972; to be published in conference conducting can then be correlated with proceedings (LBL-845). theoretical estimates. The scope of the measurements is dependent on the availability *7. R. Dalven, Possible Inverted Band Struc of suitable samples. ture in PbPo, Phys. Rev. Letters 28, 91 (1972). The measurements of l/i noise in thin films of bismuth and gold will be completed. *8. R. Dalvea, Electronic Structure of PbS, We hope to be able to correlate the noise PbSe, and PbTe, in Solid State Physics, with the structure of the films as deter edited by F. Seltz, D. Turnbull, and mined by electron microscopy. We also plan H. Ehrenreich (Academic Press), Vol. 28 to perform further computer calculations on (to be published 1973). our nonlinear model. *9. R. Driven, Electron and Hole Effective A project will be initiated to construct Masses in Lead Polonide (PbPo), J. Phys. C: a thin film gradiometer, using a dc SQUID Solid State Phys (to be published 1973). and our resonant circuit readout technique. It is hoped that this device will be very R. Dalven, Calculated ilnergy Gap ard stable, and reliable enough to be used in a Effective Masses in Lead Polonide (submitted non-laboratory environment. to Phys. Rev.). We plan to start measurements on the Conferences resistance of very clean Pb-Cu-Pb sandwiches. We hope to see abrupt changes in resistance 1. J. Clarke, S. M. Freake, and M. L. as a function of temperature, arising from Rappaport, Measurement of Very Low Super- -204- conducting Transition Temperature.;, San 3. J. Clarke, Observation of Pair- Francesco Meeting cf the American Physical Quasiparticle Potential Differei.ces in Non- Society, J-3nuary 31-Feburary 3, .'972; Bull. equilibrium Superconductors, paper at Am. Phys. Soc. 17, 46 (1972) flBi-423 Abs.). Thirteenth International Conference on Low Temperature Physics, Boulder, Colorado, August 21-25, 1972 rr,T'L-855 Abs.). 2. J. Clarke, Macroscopic Quantum Devices, invited paper at Thirteenth International Conference on Low Temperature Physics, Boulder, Colorado, August 21-25, 1972 R. Dalven, presently with the Dept. of (LBL-1144 Abs.). Physics, U. C. Berkeley and a participant with IMRD. -205- D. EXPER IMENTA!. SOLID STATE PHYSICS AND QUANTUM ELECTRONICS J. Ron Shen, Principal Investigator the similarity of their molecular structure. A few Raman modes, particularly the inter- 1. RESONANCE RAMAN SCATTERING IN InSb molecular ones, exhibit quasidiscontinuous changes at the phase transitions. These Peter Y. Yu and Y. Ron Shen changes are qualitatively interpreted as the result of i edification of the inter- Resonance Raman scattering in InSb near the molecular interaction induced by the phase Ex transition has been measured at low transitions. temperatures for various different configura tions with a tunable dye laser. The experi Vfe have investigated the Raman spectra of mental curves show a resonance peak at an sonis smectic compounds in the spectral region incident photon energy appreciably higher of 5 to 2000 cm"1. Emphasis was given to the than the energy of the Ei peak in the reflec terroerature dependence of the low frequency tivity spectra, contrary to the predictions modes near the phase transitions. We observed of existing theories. Double resonance in a low frequency mode in the solid phase of the Raman tensor involving transitions in the the related compounds diethylazoxybenzoate band continuum is proposed to explain the (DEAB) and diethyazoxycinnamate (DEAC). This result. mode shifts abruptly to lower frequency and decreases in intensity at the transition to smectic while vanishing in the isotropic phase. Since the frequency of this mode is 2. RESONANCE RAMAN SCATTERING AND found to be independent of the length of the LUMINESCENCE IN Cu20 end groups, we interpret it as arising from the intermolecular dipole-dipole interaction Peter Y. Yu, Yves P. Fetroff, and Y. Ron Shen between the carboxyl groups. The shift in the frequency of this mode is attributed to Luminescence spectra of Cu20 at various increase in the spacing of the layered temperatures excited by a tunable dye laser structure of the smectic phase. This is con showed no evidence of the existence of sistent with the results obtained from X-ray. excitonic molecules (or biexcitons) as The same interpretation explains the decrease suggested recently by Gross et al. A pair in intensit/ of this mode. No low frequency of previously unreported peaks with an modes were observed in other smectic com exceptionally strong temperature dependence pounds such as the higher homologues of has been observed below 3°K. axoxybenzene, presumably because of the difference in the intermolecular interaction. We also obtained strong resonance enhance ment in the 220 cm"1 Raman line of CU2O at the r12 -phonon-assisted threshold of the forbid den Is yellow exciton. A consistent theoreti 4. RAYLEIGH-WING SCATTERING NEAR THE NEHATIC- cal interpretation is found. It shows that ISOTROPIC TRANSITION the line is due to scattering of two rj2 phonons and the resonance enhancement is Nabil M. Amer, David Y. S. Lin, and strongly affected by the exciton lifetime. Y. Ron Shen The latter effect has never been considered or observed before in resonance Raman WhiJe fluctuations in tiie direction of scattering. molecular alignment can give rise to ^ijht scattering in a liquid crystal , fluctuations in the orientation of individual molecules can also cause light scattering. The latter 3. RAMAN AND RAYLEIGH SCATTERING IN LIQUID is usually described as Rayleigh-wing scat CRYSiALS tering. We have investigated for the first time the Rayleigh-wing scattering in a nematic Nabil M. Amer, David Y. S. Lin, and Y. Ron compound MBBA in both the nematic and the Shen isotropic phases as a function of temperature. Our results indie-te that the spectrum of We have investigated the Raman spectra of the Raylei^h wing can be decomposed into two four nematic members of the homologous series Loientziw, a narrow one and 1 broad one. 4, 4'-bis (alkoxy) azoxybenzene in the solid, As the temperature approaches the isotropic- nematic, and liquid phases for r. spectral nematic transition from the isotropic side, range of 10 to 1900 cm"l. These spectra show both Lorentzians become narrower, indicating great similarity, as one would expect from a slowing down in the orientational fluctua- -206- tions. The phenomenon is somewhat similar 7. SECOND-HARMONIC GENERATION FROM A SOLID- to that of spin fluctuations in a magnetic STATE PLASMA. system near the order-disorder transition. Arnold Schmidt, Eric G. Hanson, and Y. Ron Shen Plasmons in metals have been excited by a 5. WAVELENGTH-MODULATION SPECTROSCOPY OF number of different processes. One conceiva SOLIDS ble process is the generation of a plasircm by a two-photon excitation. A plasmon Yves P. Petroff, Stanley E. Kohn, decaying radiatively will emit photons of Jeffrey A. Stokes and Y. Ron Shen twice the frequency, which might be detected. Silver is a suitable candidate for such an We have used our wavelength-modulation experiment since its plasmon frequency spectrometer to study optical properties of (~3300 A) is in reach of available tunable semiconductors and metals. The high sensi ciye lasers. Attempts have been made to tivity and high resolution of the spectrom obtain a sufficiently powerful laser at eter enables us to find many new features in 6600 A. the solid spectra. The spectrometer has also been used to study the changes in the Cresyl violet has been investigated, either spectra induced by external perturbations. pumped with frequency-doubled light from a Our results on InP, GaP, ZnGeP2, SbSI, Ag, Nd laser or by a flash lamp. Both methods Cu, and Au have been obtained and compared have led to lasers which might be powerful with the spectra calculated by Marvin L. Cohen enough for the planned experiment. and his associates using the empirical pseudopotential method. The agreement be tween theory and experiment is very good. 8. FAR INFRARED GENERATION BY OPTICAL MIXING 6. DISPERSION OF NONLINEAR OPTICAL James R. Morris, Patrick K. H. Yang, SUSCEPTIBILITIES Paul L. Richards, and Y. Ron Shen Arnold Schmidt, Don Bethune, and Y. Ron Shen We have built a novel dye laser cavity which allows two orthogonally polarised, separately The disoersion of the nonlinear suscepti tunable laser beams to be generated bility \"i close and above the band gap of simultaneously in one cavity. Since both semiconductors is of considerable interest. beams are generated in a common pumped region and utilize the same output mirror, A setup to measure nonlinear susceptibili the problem of aligning the beams for some ties against a known reference (e.g., Quartz) dual beam nonlinear optical spectroscopy was developed. The fundamental frequency experiments including far infrared generation is provided by a dye laser pumped by ruby by optical mixing is greatly simplified. We light or its second harmonic, or alternatively have used this system in preliminary by a flash lamp pumped dye laser. The investigations of LiNb03, ZnO, ZnS, CdS, spectral region covered exte.nds from 4100- and CdSe as possible far infrared sources 9000 A. The detection system employs gated in the region 10 to 250 cm"l. LiNb03 seems electronic integrators. The first investiga to be the best candidate for the low tions will be on III-V compounds above their frequency range, say below 50 cm"l because of band gap. Measurements on InAs have been its high electro-optic coefficient. ZnO started. seems to be the best candidate around 180 cm"l due to its lower absorption coefficient. Another experiment planned is the measure CdS may be useful in the frequency range ment of the nonlinear susceptibility of below 180 cm"l also. We are also using the alkali halides. Since these materials have nonlinear mixing process to directly measure a center of symmetry, the nonlinear the absorption coefficient of Li*. .1)03 *" a susceptibility is very 'imall. However, at region inaccessible to ordinary far infrared the exciton frequency r. large resonance techniques and to measure the dispersion of enhancement is to be expected. the lowest order nonlinear susceptibility. -207- 9. COHERENT PHONON GENERATION BY OPTICAL the spectrometer further into the vacuum UV. MIXING We shall conduct measurements on the dis persion of nonlinear optical susceptibilities Patrick K. H. Yang, Paul L. Richards, and in many nonlinear media. We also plan to Y. Ron Shen spend more effort in studying the linear and nonlinear optical properties of solids. We have examined the feasibility of generating high frequency coherent phonons in a superlattice by optical means. With the help of the lattice momentum, phase- 12. 1972 PUBLICATIONS AND REPORTS matched optical generation of coherent phonons with frequencies higher than 50 GHz appears Y. Ron Shen and Associates feasible. We have considered two cases theoretically: 1) direct conversion of Journals millimeter or submillimeter photons in a piezoelectric superlattice, and 2) optical 1. Carmen Varea de Alvarez, John P. Walter, mixing of two laser beams in a non- Marvin L. Cohen, J. Stokes, and Y. R. Shen, piezoelectric superlattice. Wavelength-Modulation Spectra and Band Structures of InP and GaP, Phys. Rev. B 6, 1412 (1972)(LBL-479). 10. SPECTRAL BROADENING IN THE SELF-FOCUSED 2. Peter Y. Yu and Y. R. Shen, Resonance LIGHT Raman Scattering in InSb Near the Ej Trans ition, Phys. Rev. Letters i.3, 468 (1972) George K. L. Wong and Y. Ron Shen (LBL-833). It has been shown recently that, at least 3. George K. L. Wong and Y. R. Shen, Study for nanosecond pulse excitation, moving foci of Spectral Broadening in a Filament of are responsible for the observation of Light, Appl. Phys. Letters 21_, 163 (1972) small-scale filaments in Kerr liquids. Most (LBL-862). observed characteristics of the filaments can be explained by the moving-focus model. 4. H. Rosen and Y. R. Shen, Brillouin Scat In particular, spectral broadening of light tering in a Cholesteric Liquid Crystal Near from a filament can be interpreted as due the Cholesteric-Isotropic Transition, to phase modulation acquired by the self- Molecular Crystals and Liquid Crystals 18, focused light in traversing the nonlinear 285 (1972)(LBL-452). ~ medium. However, no controlled systematic experiment on spectral broadening has yet 5. Nabil M. Amer and Y. R. Shen, Raman Scat been performed to check the validity of the tering from Nematic Liquid-Crystalline interpretation. Such an experiment is im Azoxybenzenes, J. Chem. Phys. 56, 2654 (1972) portant in view of the fact that spectral (UCRL-20564). broadening has long been used as evidence to support the self-trapping model. We have 6. J. W. Shelton and Y. R. Shen, Study of performed the first experiment on spectral Phase-Matched Normal ••md Uniklapp Third- broadening with nanosecond input pulses under Harmonic-Generation Processes in Cholesteric controlled input conditions. We show that Liquid Crystals, Phys. Rev. A 5, 1867 (1972) there exists, in fact, semiquantitative (LBL-161). agreement between theory and experiment. Induced partial trapping or nonlinear dif 7. J. Stokes, Y. R. Shen, Y. W. Tsang, fraction of light in a filament can occur, M. L. Cohen, and C. Y. Fong, Wavelength but its effect is only on the detailed Modulation Spectra of Single Crystals of structure of the broadened spectrum. Silver and Gold, Phys. Letters 38A, 347 (1972)(LBL-451). 8. Y. Petroff, P. Y. Yu and Y. R. Shen, 11. RESEARCH PLANS FOR CALENDAR YEAR 1973 Luminescence in Cu20--Exciton Molecules or Not, Phys. Rev. Letters 29, 1558 (1972) Y. Ron Shen (LBL-1133). We shall continue our research on resonance Raman scattering and luminescence in several 9. J. Koo, Y. R. Shen, and R. R. L. Zucca, different crystals. We shall also continue Effects of Uniaxial Stress on the EA-Peak our work on wavelength-modulation spectroscopy of Silicon, Sol. State Comnun. 9_, 2229 (1971). and hope to extend the operating range of (Not listed last year) (LBL-146). -208- Papers presented LBL reports 1. S. Kohn, Y. Petroff, P. Y. Yu and 1. S. Kohn, Y. Petroff, Y. R. Shen, and Y. R. Shen, Wavelength Modulation Spectra P. Yu, Wavelength Modulation Spectra of PbS, of PbS, PbSe, PbTe, and SnSe, Conference on PbSe, PbTe, and SnSe, LBL-4S6, April 1972. the Physics of IV-VT Compounds and Alloys, Philadelphia, Pa., March 1972 CLBL-456). 2. K. H. Yang, P. L. Richards, and Y. R. Shen, Coherent Phonon Generation by Optical 2. George K. L. Wong and Y. R. Shen, Study Mixing in a One-Dimensional Superlattice, of Spectral Broadening of Light in a Filament LBL-889, June 1972. Induced by a Moving Focus, Seventh Interna tional Quantum Electronics Conference, 3. S. E. Kohn, P. Y. Yu, Y. Petroff, Toronto, Canada, May 1972 (LBL-468 Abstract). Y. R. Shen, Y. Tsang, and M. L. Cohen, Electronic Band Structure and Optical Pro 3. Nabil M. Amer, Y. S. Lin, and Y. R. Shen, perties of PbTe, PbSe, and PbS, LBL-1132, Raman Study of Some Smectic Liquid Crystals August 1972. Near Phase Transitions, Fourth International Liquid Crystal Conference, Kent State Univer 4. Nabil M. Amer and Y. R. Shen, Low sity, Kent, Ohio, August 1972 (LBL-836 Frequency Raman Mode in Smectic Liquid Abstract). Crystals Near the Phase Transitions, LBL-1156, September 1972. 4. Nabil M. Amer, Y. S. Lin, and Y. S. Shen, Rayleigh-Wing Scattering Near the Nematic- 5. Ricardo Zucca and Y. R. Shen, A Wide Isotropic Transition, Fourth International Range Wavelength Modulation Spectrometer, Liquid Crystal Conference, Kent State LBL-1184, October 1972. University, Kent, Ohio, august 1972 (LBL-835 Abstract). 6. P. Y. Yu, Y. R. Shen, Y. Petroff, and L. M. Falicov, Resonance Raman Scattering at 5. Y. R. Shen, Optica? Wavelength-Modulation the Forbidden Yellow Exciton in CujO, Spectroscopy, Invited talk presented at the LBL-1192, October 1972. First International Conference on Modulation Spectroscopy, Tucson, Arizona, November 1972 7. Y. R. Shen, Optical Wavelength-Modulation (LBL-1402). Spectroscopy, LBL-1402, November 1972. 6. S. Kohn, Y. Petroff, and Y. R. Shen, 8. S. Kohn, Y. Petroff, and Y. R. Shen, Wavelength Modulation Spectra of GaSe, GaS, Wavelength-Modulation Spectra of GaSe, GaS, and Alloys, First International Conference and Alloys, LBL-1407, November 1972. on Modulation Spectroscopy, Tucson, Arizona, November 1972 (LBL-1407). 9. Y. Petroff, S. Kohn, and Y. R. Shen, Optical SpectTa of CSI, CsBr, ZnGeP2, and 7. Y. Petroff, S. Kohn, and Y. R. Shen, SbSI, LBL-1409, November, 1972. Optical Spectra of Csl, CsBr, ZnGeP2, and SbSI, First International Conference on 10. Y. Petroff, S. Kohn, and Y. R. Shen, Modulation Spectroscopy, Tucson, Arizona, Effect of Carrier Concentration on the November 1972 (LBL-1409). Reflectivity Spectrum of Ge and InSb, LBL-1410, November 1972. 8. Y. Petroff, S. Kohn, and Y. R. Shen, Effect of Carrier Concentration on the Reflectivity Spectrum of Ge and InSb, First International Conference on Modulation Spectroscopy, Tucson, Arizona, November 1972 (LBL-1410). -211- IMRD STAFF FOR 1972 Principal Postdoctoral Fellows Investigators and Other Scientists Graduate Students Neil Bartlett *C. Adams C. Kindle *E. Moraga K. Leary *W. Roland D. McKee *A. Tressaud T. Richardson *B. Zemva Robert Bragg K. Antra M. Bisual *Y. Chen K. Danso F. Hamzeh L. Salmon R. Saxena Leo Brewer S. Chang *G. Bullard P. Desai S. Chang D. Jensen R. Macklin *G. Rosenblatt J. Stiga P. Yee John Clarke *R. Dalvtii G. Hoffer S. Freake *M. Ketchen J. Kruger J. Paterson M. Rappaport L. Schwartzkopf R. Voss Marvin Cohen *L. Saravia *R. Boyd *D. Chadi *J. Chelikowsky *J. Joannopolous *S. Sramek *C. Varea de Alvarez Robert Connick W. Earl T. Rowland Lee Donaghey C. Wright R. Agnihotri R. Bissinger *P. Chen *P. Cross R. Headrick *A. Kaya P. Mundkur R. Pong S. Shaikh Richard Fulrath G. Bandyopadhyay D. Biswas K. Crabtree J. Dih R. Holman J. Masaryk J. Sherohman C. Shumaker D. Wang -212- Principal Postdoctoral Fellows Investigator and Other Scientists Graduate Students Charles Harris *H. Brenner W. Breiland *T. Gough *J. Brock R. Hoover M. Fayer A. Nishinura W. Huglies R. Schlupp J. Koert *H. Schuch R. Panos *R. Shelby *J. Weber P. Wilday Ronald Hera R. Behrens B. Earl A. Freedman L. Gundel S. Lin L. Loh C. Minis T. Parr C. Sholeen A. Wong Ralph Hultgren J. Cases P. Desai K. Kelley HaTold Johnston S. Chang J. Birks *D. Fielder E. Franzblau *E. Kamaratos R. Graham T. Paukert A. Harker G. Whitten D. Martin M. Moran P. Patnaik *G. Streit *J. Valdin R. Young William Jolly P. Finn *R. Garbutt J. Howdeshell J. Illige C. Riddle M. Mendelsohn *J. Selbin W. Perry L. Pont J. Reed K. Strom George Jura T. Chen R. Glienna B. Jurado Bruce Mahan *K. Gillen W. Dinpfl T. Sloane J. Fair P. Schubart *R. Terkowitz J. Winn -213- Principal Postdoctoral Fellows Investigators and Other Scientists Graduate Students William Miller *J. Doll S. Augustin T. George B. Garrison *W. Hunt S. Hornstein *R. Preston A. Raczkowski *C. Sloane *P. Whitlock John Morris, Jr. D. Klahn I. Charo K. Murty S. Dong D. Grivas K. Hanson S. Jin D. Klahn C. Krause F. Mohamed C. Syn Rolf Muller *J. Mathieu C. Brown J. Cooper F. McLarnon W. Smyth Rollie Myers T. Hynes M. St. John R. Wilson J. Yarnell John Newnan P. Appel R. Homsy C. Mohr K. Nisancioglu N. Vahdat H. Yip Donald Olander A. Oilman R. Acharya *C. Cann V. Kruger R. Lu A. Machiels *G. Moore *F. Nuh R. Olstad *D. Slaughter H. Tsai Earl Parker M. Collver B. Babu Victor Zackay *G. Lai B. Francis Milton Pickus M. Yokota K. Hemachalam T. Tom S. Wason -214- Principal Postdoctoral Fellows Investigators and Other Scientists Graduate Students Earl Parker *R. McCoy H. Adkins Victor Zackay D. Atteridge A. Baghdasarlir. D. Bhandarkar M. Bhat S. Dong J. Ellis C. Ericsson S. Friedrich D. Griffiths W. Harris P. Horn w. Hnnsiod R. Lott H. Offer W. Oliver M. Perra E. Plaza G. Sasaki E. Saur M. Schanfein H. Sirot W. Wood P. Zajchowski Joseph Pask R. Langston I. Aksay C. Dokko C. Hoge W. Sncwden T. Sweeting C. Tao B. Wong Norman Phillips M. Conway S. Bader M. Conway W. Fogle T. Hall G. Schwartz Milton Pickus L. Miwa Alan Pcrtis Paul Richards *S. Buckner R. Aurbach J. Claassen *J. Mather W. Marzouk Y. Taur *M. Werner W. Tennant D. Woody K. Yang Gene Rochlin P. Hansma P. Hansma M. Jack D. McBride Alan Searcy A. Buchler P. Bihuniak D. Meschi D. Feather L. Fuke P. Mohazzabi J. Roberts -215- Principal Postdoctoral Fellows Investigators and Other Scientists Graduate Students Y. Ron Shen Y. Petroff A. Bromberg *A. Schmidt W. Garrison P. Yu E. Hanson S. Kohn I. Ku Y. Lin J. Morris W. Nevins J. Stokes K. Wong Gabor Somorjai K. Baron S. Bernasek *S. Berglund D. Blakely *J. Deville *L. Fitment B. Lang J. Gland M. Martin D. Kahn *W. Siekhaus D. Megerle *L. West R. Palmeri F. Szalkowski J. Wasilczyk Gareth Thomas W. Bell L. Chen M. Bouchard R. Clark *W. Mailer R. Gronsky M. Raghavan R. Hall *K. Schneider D. Howitt *M. Von Heimendahl D. Huang R. Livak J. McMahon R. Miller F. Mimiirani M. Okada M. Raghavan G. Stone 0. Van der Biest Charles Tobias *T. Lengas R. Acosta L. Silvester *E. Carlson J. Jorne1 K. Kojima U. Landau M. Sand Jack Washburn J. Narayan S. Alavi T. Astrup R. Jecmen T. Kosel J. Lira-Olivares H. Mohamed T. Peterson 3. Roberts K. Seshan A. Srivastava K. Stolt W. Wu •Participant Status -216- VISITING SCIENTISTS Dr. A. S. Abbasov, Azerbaidzhan Academy of Sciences, U.S.S.R. Professor George Benedek, Department of Physics, Massachusetts Institute of Technology, Cartridge, Massachusetts Dr. Robert Berg, Department of Chemistry, Technion-Israel Institute of Technology, Haifa, Israel Dr. Igor Borovskii, Laboratory Chief and Doctor of Physico-Mathematical Sciences at the A. A. Baykov Institute of Metallurgy, Moscow, U.S.S.R. Dr. Spencer Bush, Battelle Memorial Institute, Pacific Northwest Laboratory, Richland, Washington Dr. Leonard Dobrzynski, Department of Physics, University of California, Irvine Professor P. H. Emmett, John Hopkins University, Baltimore, Maryland Dr. Daniel Fiat, Weizmann Institute of Science, Rehvot, Israel Professor Heinz Gerischer, Director of the Fritz Harber Institute, West Berlin, Germany Professor Robert F. Hehemann, Department of Material Science and Engineering, Case Western University, Cleveland, Chio Dr. H. E. Hintermann, Head, Physical Chemistry Section, Swiss Watch Research Laboratory, Neuchatel, Switzerland Professor W. Hirschwald, Free University of Berlin, Germany Dr. James Hoffer, Los Alamos Scientific Laboratory, Los Alamos, New Mexico Dr. A. Ludi, Institute for Inorganic, Analytic and Physical Chemistry, University of Bern, Bern, Switzerland Frofessor S. F. Mason, Department of Chemistry, University of East Anglia, Norwich, England Professor N. Rambidi, Institute for High Temperature, Moscow, U.S.S.R. Professor E. Saur, Director of Institute at the University of Giessen, Giessen, Germany Professor M. J. Sienko, Department of Chemistry, Cornell University, New York Dr. William E. Silverthom, Department of Inorganic Chemistry, University of Oxford, Oxford, England Dr. Y. Song, Department of Physics, Simon Fraser University, Bumaby, British Columbia, Canada -217- DEGREES AWARDED IN 1972 Ph.D. Degrees 1. Charles J. Bruggemann, The Effect of One-Dimen?"->rial Plastic Strain Waves in a-Phase Titanium- Aluminum Alloys (I.BL-802, April). 2. Shih-Ger Chang, High Temperature Spectroscopy and Thermodynamics of Titanium Oxide and Vanadium Oxide (LBL-450, March). 3. Mervyn Ming-Hwa Chiang, Reactive and Inelastic Scatterings of Molecular Ions (LBL-474, February). 4. Mary Mclvor Conway, The Low-Temperature Heat Capacity of Cerium Metal (LBL-827, May). 5. Roy Tom Coyle, Vaporization Kinetics of Magnesium Nitride (LBL-136, January). 6. David Hoover Feather, Mass Spectrometric Studies of Inorganic Vapors (LBL-844, September). 7. Paul Kenneth Hansma, Externally Shunted Josephson Junctions (LBL-498, March). 8. Alan Butler Marker, Reaction Kinetics of the Photolysis of NO2, and the Spectra and Reaction Kinetics of the Free Radicals in the Photolysis of the CI2-O2-CO System (LBL-1114, September). 9. Robert Louis Holman, Intrinsic and Extrinsic Non-Stoichiometry in the Lead Zirconate-Titanate System (LBL-880, August). 10. Jacob JcmS, Electrochemical Behavior of the Alkali Metals in Propylene Carbonate CLBL-1111, September). 11. Dale Henry Klaki, Dislocation Modeling of Strain-Rate Phenomena (LBL-800, March). 12. Kaoru Kojima, A Study of the Mechanism of Electropolishing of Copper (LBL-140, February). 13. Shen-Maw Lin, Molecular Beam Kinetics (LBL-891, September). 14. Ronald John Livak, Spinodal Decomposition and Coarsening in Cu-Ni-Fe Alloys CLBL-1107, September). 15. Marshall Harvey Mendelsohn, Reactions of Heptasulfur Imide and the Heptasulfur Imide Anion (LBL-495, March). 16. Farghalli Abdel Rahman Mohamed, High Temperature Creep in Metals at Intermediate and Low Stresses (LBL-1183, November). 17. Robert Allan Olstad, Mass Spectrometric Investigation of Laser-Induced Vaporization of Binary Solid Compounds (LEL-1177, December). 18. Kenneth Alan Strom, Reactions of Hydrogen in Liquid Ammonia Systems (LBL-1408, November). 19. Teh Yu Tan, Behaviors of Inelastic Electrons (LBL-432, January). 20. Francis Ting-yuan Wang, A Kinetic Study of the Stepwise Hydrolyses of the Hydmborate and the Octahydrotriborate (LBL-411, February) . D.Eng. Degrees 1. Mathur Ramachandran Veera Raghavan, Structure and Mechanical Properties of Fe-Ni-Co Alloys with and without Carbon (LBL-477, March). M.S. Degrees 1. Rajani Bhushan Agnihotri, Chemical Transport Reaction Analysis for C-ystal Growth of Binary and Ternary Chalcogenides (LBL-1404, December). 2. Kwame Ankra, Unidirectional Solidification of AI-CUAIT Eutectic (LBL-1182, December), 3. Ara Joseph Baghdasarian, Corrosion Resistance of TRIP Steels (LBL-806, May). 4. Peter Paul Bihuniak, The Mechanism of Zinc Oxide Single Crystal Vaporization (LBL-867, June). 5. Richard WyantHeadrick, Design Criteria for Solid Electrolyte Electrochemical Cells (LBL-839, July). -218- 6. Kanithi Hemachalam, Evaluation of Optimal Process Parameters for the Production of Super conducting Niobium-Tin in Relation to Pilot Plant Design (LBL-1108, June). 7. Wade Alan Horwood, Factors Controlling the Fracture Toughness of Cryogenic Alloys (LBL-1121, December). 8. John Arthur McMahon, Structure, Strength and Toughness of Fe/Cr/C Martensitic Steels (LBL-1181, December). 9. Pirooz Mohazzabi, The Kinetics of Vaporization of Barium Sulfate (LBL-1416, December). 10. Mark Walter Perra, Nuclear Magnetic Resonance Study of GP Zone Formation in Al-4%Cu (LBL-838, June). 11. Saleem Anwar Shaikh, Chemical Vapor Deposition of GaAsj.xPx Reactor Design and Growth Kinetics (LBL-1149, September). 12. Craig Barlow Shuma'.cer, Initial Stage Sintering of Copper and Nickel (LBL-1110, August). 13. Chi-Shine Cliff Tao, Polycrystalline Spinel from Powders Prepared by Freeze Drying Technique (LBL-876, June). 14. Nader Vahdat, The Corrosion of Iron Rotating Disks (LBL-873, June). 15. Sarwan Kumar Wason, Screening of Candidate Materials for Bio-Compatibility (LBL-None, March). 16. Boon Wong, Permeation of Silicate Liquids into Sintered Magnesia Compacts (LBL-877, September), -219- 1972 RESEARCH SEMINARS Date Title Speaker 2/17 The Stress-Induced Martensitic Transformation of Glen A. Stone (GS)* Fe-Ni-Cr Austenitic Stainless Steel Single Crystals 2/17 Intrinsic Non-Stoichiometry in the Lead Titanate-Lead Robert L. Holman (GS)* Zirconate System 3/9 Kinetics of Vaporization of Zinc Oxide Professor W. Hirschwald Free University of Berlin Germany 3/9 Effect of Martensitic Substructure on Mechanical Mathur Raghavan (GS)* Properties 3/9 Ordered Arrays of Stops on High. Index Faces of Platinum Bernard Lang (PD)t Crystals 3/9 The Resistance of TRIP Steels to Hydrogen Embrittlement Robert A. McCoy (PD)+ (H.E.) 3/21 Low Temperature Magnetic Studies of Metal-Ammonia M. J. Sienko Compounds Cornell University Department of Chemistry New York 3/23 Antiphase Domains in Terrestrial and Lunar Anorthite Wolfgang MuLler (PD)+ 3/23 A Study of the Mechanism of Electropolishing of Copper Kaoru Kojima (GS)* 3/23 Current Gain in Josephson Junctions James L. Paterson (GS)* 4/6 On the Stability of Dislocation Loops Near Free Surfaces Jagdish Narayan (PD)''' 4/6 Correlation of Thermo-Mechanically Varied Microstructures B. Naga Babu (GS)* with Superconducting Properties of l*3Sn Processed by a New Powder Metallurgy Technique 4/6 Low Temperature Heat Capacity of Cerium Metal Mary I. Conway (GS)* 4/20 Anodic Behavior of Titanium Carbide in Sulfuric Acid Dr. H. E. Hintermann Swiss Watch Research Laboratory Neuchatel, Switzerland 5/31 Surfaces, Relaxation, Reconstruction, Adsorption Leonard Dobrzynski Department of Physics University of California Irvine 7/6 Dislocation Motion Through a Random Array of Point Obstacles Dale Klahn (GS)* 7/6 Electrochemical Behavior of the Alkali Metals in Propylene Carbonate Jacob Jome" (GS)* 7/10 Structural Chemistry of Polynuclear Transition [-fetal A. Ludi Cyanides Universitat Bern Switzerland -220- Date Title Speaker 8/25 Behavior of Type II Superconductors with AOCV) and Professor E. Saur B(I) Structure in High Magnetic Fields Director of the Institute of the Angewandte Physik University of Giessen Germany 10/19 Spinodal Decomposition in Cu-Mn-Al Alloys Marcel Bouchard (PD)+ 10/19 The Oscillator Strength of Ti-0 oc System and TO Green Shih-Ger Chang (GS)* System 10/19 A Systematic Approach to the Band Structure of III-V Jeffrey Stokes [GS)* Semiconductors 11/30 Mass Spectrometric Investigation of Laser-Induced Robert Olstad (GS)* Vaporization of Binary Solid Compounds 11/30 A Method for Increasing the Fracture Toughness of William Wood (GS)* Ultra-High Strength Steels 11/30 Surface Features in the Evaporation of As Single Crystals Dr. G. Rosenblatt (C) Peim State 12/7 Use of Ceramics as Surgical Implants Ralph LeonaH Clemson University South Carolina Graduate Student tPostdoctoral Fellow Consultant